^^^ 
 
 'bV'' 
 
 
 < V "^ * o • » A' 
 
 
 .•^°^ 
 
 
 <&■ ,.'"'<', 
 
 -Jv^ , 
 
 .^ 
 
 
 
 .-l^^ 
 
 
 <> *; 
 
 A 
 
 
 
 ,> 
 
 
 \/ 
 
 .*1°<. 
 
 
 ■^-. .^^^^i-^'-"^ .^^^•^i'% ^'^:^^'> .^^ -^i^^ /.-•- -^ 
 
 
 
 > 
 
 ,v ., V'^'^'*/ V'*^^^\^''' v^*"-'^ 
 
 .^'\ 
 
 
 •>o^ 
 
 
 * '^ 
 
 •^6" 
 
 
 v-'-y V'3'\/\_ V'"^"'/' "v'^'^V V'"^^'"/' '\ **'•■■• A^ 
 
 *-'-i--\ «<•«'> /--^i-X o'°-]e^-> /--^i-X .^""•- 
 
 4 CU 
 
 
 .0^ ..-4*,**; 
 
 
 ^^ 
 
 '^_ 
 
 A , « o . <i> 
 
 ^^ ...*. -tU 
 
 
 •^-^ 
 
 .(X.* -o»a. <S 
 
^^ r^"" t^ ''T.^* A <*. '*?5!^\6^ 'o 
 
 
 <^^ , « o 
 
 ^'« ^<tp / *;<^lfA^o u '^ ^^^M' \ / /-A^^vo ^.. .^ .^1 
 
 
 9 < 
 
 
 
 
 
 .•^°<. 
 
 'btv" 
 
 
 
 ' . . s 
 
 \^^^ :^^M\ %/" y^g° %^^^ ^^^fe°' "-^'^ 
 
 0- "V, * " ■• * . ■?> 
 
 -~ .n 
 
 
 • .^^ 
 
 cCl°^ 
 
 ^^-^^^ 
 
 > _ 5 • • . •>. (jV ^ , , , 
 
 
 
 
 
 
 .-^^^ 
 
 
 '-^"^-^ .'^".^i'^^'^^ ^^*'^•^t^''-^ ^''^^^%''^ .^^^*^<'''''- o^\^"-'/^' ' 
 
 
 0' 
 
 1^7 o C, vP 
 
 
 
 A <A *v 
 
 
 
 * AT ^ 
 » . »1* IX* a X o -<> 
 
IC 
 
 9059 
 
 Bureau of Mines Information Circular/1986 
 
 Precious Metals Recovery From Low- Grade 
 Resources 
 
 Proceedings: Bureau of Mines Open Industry Briefing Session 
 at the National Western Mining Conference, Denver, CO, 
 February 12, 1986 
 
 Compiled by Staff, Bureau of Mines 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 
Information Circular 9059 
 
 Precious Metals Recovery From Low- Grade 
 Resources 
 
 Proceedings: Bureau of Mines Open Industry Briefing Session 
 at the National Western Mining Conference, Denver, CO, 
 February 12, 1986 
 
 Compiled by Staff, Bureau of Mines 
 
 UNITED STATES DEPARTMENT OF THE INTERIOR 
 Donald Paul Model, Secretary 
 
 BUREAU OF MINES 
 Robert C. Norton, Director 
 

 Library of Congress Cataloging in Publication Data: 
 
 Precious metals recovery from low-grade resources. 
 
 (Bureau of Mines information circular ; 9059) 
 
 Includes Bibliographies. 
 
 Supt. of Docs, no.: I 28.27:9059. 
 
 1. Precious metals— Metallurgy— Congresses. I. United States. 
 Bureau of Mines. II. National Western Mining Conference ( 1986 : Den- 
 ver, Colo.). III. Series: Information circular (United States, Bureau 
 of Mines) ; 9059. 
 
 <SN295.U4 [TN7591 622s [669'.2l 85-600302 
 
fe 
 
 o 
 
 PREFACE 
 
 i 
 
 Q- 
 
 Most of the papers included in this Information Circular were present- 
 ed at a Bureau of Mines open industry briefing held in conjunction with 
 the National Western Mining Conference on February 12, 1986, in Denver, 
 ^ CO. The Bureau often sponsors meetings of this nature in an effort to 
 
 move new technology into industry practice by drawing attention to de- 
 velopments that may solve certain problems or improve upon current tech- 
 niques. Those desiring more information about Bureau research programs 
 should contact the Bureau of Mines, Branch of Technology Transfer, 2401 
 E Street, NW, Washington, DC 20241. 
 
 
 3 
 
 3 
 
Ill 
 
 CONTENTS 
 
 Page 
 
 Preface 1 
 
 Abs tract 1 
 
 Introduction. 1 
 
 Ion-Exchange Research in Precious Metals Recovery, by Glenn R. Palmer 2 
 
 Staged Heap Leaching-Direct Electrowinning, by C. H. Elges and 
 
 M. D. Wroblewski 10 
 
 Mercury Precipitation During Cyanide Leaching of Gold Ores, 
 
 by Richard G. Sandberg 19 
 
 Carbonaceous Gold Ores, by B. J. Scheiner 26 
 
 Carbon Adsorption-Desorption, by J, A, Eisele 34 
 
 Heap Leaching, by J. A. Eisele 37 
 
 The Carbon-in-Pulp Process, By Stephen D. Hill 40 
 
 Precious Metals Recovery From Electronic Scrap and Solder Used in Electronics 
 
 Manufacture, by B. W. Dunning, Jr 44 
 

 UNIT OF MEASURE ABBREVIATIONS USED IN 
 
 THIS REPORT 
 
 A 
 
 ampere 
 
 L/min 
 
 liter per minute 
 
 A/ft2 
 
 ampere per square foot 
 
 mg 
 
 milligram 
 
 A/in2 
 
 ampere per square inch 
 
 mg/L 
 
 milligram per liter 
 
 °C 
 
 degree Celsius 
 
 mg/min 
 
 milligram per minute 
 
 cm 
 
 centimeter 
 
 mg/(min«g) milligram per minute 
 
 
 
 
 per gram 
 
 cm^ 
 
 cubic centimeter 
 
 
 
 
 
 min 
 
 minute 
 
 ft 
 
 foot 
 
 
 
 
 
 mL 
 
 milliliter 
 
 g 
 
 gram 
 
 
 
 
 
 ml./min 
 
 milliliter per minute 
 
 gal 
 
 gallon 
 
 
 
 
 
 mm 
 
 millimeter 
 
 g/cm3 
 
 gram per cubic centimeter 
 
 
 
 
 
 mt 
 
 metric ton 
 
 g/L 
 
 gram per liter 
 
 
 
 
 
 pet 
 
 percent 
 
 g/mt 
 
 gram per metric ton 
 
 
 
 
 
 ppm 
 
 part per million 
 
 h 
 
 hour 
 
 
 
 
 
 st 
 
 short ton 
 
 in 
 
 inch 
 
 
 
 
 
 st/d 
 
 short ton per day 
 
 kg 
 
 kilogram 
 
 
 
 
 
 st/wk 
 
 short ton per week 
 
 kg/mt 
 
 kilogram per metric ton 
 
 
 
 
 
 tr oz 
 
 troy ounce 
 
 kW'h/st 
 
 kilowatt hour per short 
 
 
 
 
 ton 
 
 tr oz/st 
 
 troy ounce per short 
 ton 
 
 L 
 
 liter 
 
 
 
 
 
 V 
 
 volt 
 
 lb 
 
 pound 
 
 
 
 
 
 vol 
 
 volume 
 
 lbf/in2 
 
 pound (force) per square 
 
 
 
 
 inch 
 
 wt pet 
 
 weight percent 
 
 Ib/st 
 
 pound per short ton 
 
 yr 
 
 year 
 
 lb/(sf h) 
 
 pound per short ton 
 per hour 
 
 
 
PRECIOUS METALS RECOVERY FROM LOW-GRADE RESOURCES 
 
 Proceedings: Bureau of Mines Open Industry Briefing Session 
 
 at the National Western Mining Conference, 
 
 Denver, CO, February 12, 1986 
 
 Compiled by Staff, Bureau of Mines 
 
 ABSTRACT 
 
 This information circular describes research the Bureau of Mines has 
 conducted to improve technology for recovering precious metals from low- 
 grade resources. Many of the reported findings are recent, and some of 
 the work described is ongoing. Topics discussed include cyanidation of 
 carbonaceous gold ores to enhance gold recovery, a new method for pre- 
 cipitating mercury during cyanide leaching of gold ores, a staged heap 
 leaching process to generate suitable solutions for direct electrowin- 
 ning of gold, the use of anion-exchange resins to recover gold from cya- 
 nide solutions, and precious metals recovery from electronic scrap. 
 
 INTRODUCTION 
 
 Although U.S. gold and silver resources are extensive, many of these 
 resources are extremely low in grade and cannot be developed economical- 
 ly with conventional mining and mineral processing methods. By devel- 
 oping more efficient and economical mineral recovery techniques, the 
 Bureau of Mines has helped to make these low-grade resources more acces- 
 sible. Past research by the Bureau focused on the carbon adsorption- 
 desorption, carbon-in-pulp , oxidation pretreatment , and heap leaching 
 technologies used in various phases of precious metals recovery. Cur- 
 rent research highlights even more economical and effective approaches 
 to gold and silver recovery, such as the use of ion-exchange resins, 
 staged heap leaching followed by direct electrowinning, and mercury pre- 
 cipitation from leach solutions. The results of the Bureau's research 
 in these areas, including both past and current work, are presented in 
 the following papers. 
 
ION-EXCHANGE RESEARCH IN PRECIOUS METALS RECOVERY 
 By Glenn R. Palmerl 
 
 ABSTRACT 
 
 The Bureau of Mines investigated the 
 use of ion-exchange technology in the re- 
 covery of precious metals from cyanide 
 leach solutions. Strong-base resins such 
 as Amberlite IRA-430 are nonselective, 
 and using them results in high metal 
 loading of precious metals and mercury, 
 but generally they are difficult to 
 strip. Research was conducted to devise 
 a sequential stripping technique for var- 
 ious strong-base ani on-exchange resins. 
 The sequential stripping of IRA-430, for 
 example, eluted 100 pet of the mercury 
 with 2N H2SO4, 100 pet of the silver with 
 200 g/L NaCl in IN HCl , and all 
 remaining gold with 0.75 pc 
 150 g/L plus 5 g/L NaOH. 
 
 Weak -base resins generally are selec- 
 tive and yield low metal loading, but are 
 
 of the 
 NaClO in 
 
 not difficult to strip. Research was 
 conducted to examine the possible appli- 
 cation of various weak-base anion- 
 exchange resins in the extraction of gold 
 and/or mercury from a cyanide solution. 
 Batch equilibrium experiments were per- 
 formed at different pH values to evaluate 
 which resins might show promising results 
 in an actual mill circuit. Three experi- 
 mental resins may have possible applica- 
 tion because of their effectiveness in 
 the pH range of 10 to 12. One of these 
 resins, when tested with actual mill 
 solutions in a flow-through column ex- 
 periment, loaded to about 140 tr oz/st Au 
 and produced a peak eluant concentration 
 of 35 ppm Au. 
 
 INTRODUCTION 
 
 Numerous cyanide leach carbon-in- 
 pulp (CIP) , carbon-in-column (CIC), and 
 carbon-in-leach (CIL) milling facilities 
 have been constructed to process low- 
 grade gold-silver ore deposits throughout 
 the Western United States (J_-9).2 The 
 use of activated carbon has greatly im- 
 proved the efficiency of gold recovery 
 operations. However, several drawbacks 
 are inherent with the use of activated 
 carbon: 
 
 1. Used carbon must be thermally re- 
 activated to maintain its effectiveness 
 
 M-ID • 
 
 ^Metallurgist, Salt Lake City Research 
 Center, Bureau of Mines, Salt Lake City, 
 UT. 
 
 ^Underlined numbers in parentheses re- 
 fer to items in the list of references 
 at the end of this paper. 
 
 inhibited by 
 organic re- 
 
 2. Carbon loading may be 
 the adsorption of CaC03 , 
 agents, and clay-type minerals. 
 
 3. Carbon stripping must be performed 
 in a pressurized vessel at elevated tem- 
 peratures, 120° to 130° C, to shorten the 
 stripping time. 
 
 An alternative recovery technology 
 studied by the Bureau of Mines involves 
 the application of anion-exchange resins 
 for the recovery of gold from cyanide 
 solutions. As early as 1949 (11-13) , 
 attempts were made to use both weak- and 
 strong-base anion-exchange resins. Since 
 that initial work, various applications 
 and problems associated with anion- 
 exchange resins have been examined. 
 
 Strong-base resins are generally less 
 selective for precious metals, but 
 have higher loading capacities and are 
 much less affected by pH than weak-base 
 
resins, A general mechanism for loading 
 a strong-base resin is as follows: 
 
 I-NR3+X- + M(CN)2- 
 
 ^ |-NR3+M(CN)2" + X- 
 
 (A) 
 
 where |— , NR3 , H+X" , and M(CN)2" repre- 
 sent the polymeric resin matrix, func- 
 tional group, acid, and metal-cyanide 
 complex, respectively. Weak-base resins 
 are generally more selective for precious 
 metals, but have lower loading capaci- 
 ties. Because protonation is required to 
 extract anions, a pH of less than 10 is 
 usually required (14). A general mechan- 
 ism for loading a weak-base resin may be 
 represented by the following: 
 
 I-NR2 + H+ + X^ 
 ^=> [-NR2H+X" 
 
 (Protonation) (B) 
 
 I-NR2H+X- + M(CN)2' 
 
 ^=^ |_NR2H+M(CN)2" + X- (Loading) (C) 
 
 Both resin types can be eluted at ambient 
 temperature and pressure. Weak-base res- 
 ins generally can be eluted using a di- 
 lute caustic solution. Strong-base res- 
 ins, however, are more difficult to elute 
 and require a more rigorous treatment. 
 Potassium thiosulfate, acetone plus HCl, 
 ethyl acetate plus HNO3 diluted with 
 water (13) , zinc cyanide, and dimethyl 
 formamide have each been used to recov- 
 er gold ( 15 ) . Each elution technique 
 has one or more of the following dis- 
 advantages: fire hazard, noxious gas 
 formation, or the need for special regen- 
 eration procedures. The mechanisms for 
 eluting strong- and weak-base resins are 
 described by the reverse of reactions A 
 and C, respectively. 
 
 STRONG-BASE RESINS 
 
 During the past few years, the Bureau 
 of Mines has investigated problems as- 
 sociated with the elution of gold, sil- 
 ver, and mercury from strong-base resins 
 such as Amberlite IRA-430 and 900 and 
 Dowex 21-K, SBR, and SMA-1.3 Tests were 
 initiated by loading resins with 145 tr 
 oz/st each of gold, silver, and mer- 
 cury from cyanide solutions doped with 
 radioactive tracers for analysis. The 
 loaded resins were eluted with several 
 eluants, such as H2SO4, HNO3 , acid chlo- 
 ride, hypochlorite, and/or alkaline chlo- 
 ride solutions. 
 
 Figure 1 shows a simplified flowsheet 
 for the preferred eluting sequence. Ta- 
 ble 1 presents the results of using the 
 simplified flowsheet with each of the 
 five resins. Generally, all five strong- 
 base resins behaved similarly. Within 
 the first 3 h, 100 pet of the mercury was 
 eluted with 2N H2SO4. No gold and only 
 about 10 pet of the silver was eluted 
 with the mercury; however, about 20 pet 
 
 ■^Reference to specific products does 
 not imply endorsement by the Bureau of 
 Mines. 
 
 of the silver was eluted with the mercury 
 using the SBR resin. A subsequent elu- 
 tion with 200 g/L NaCl in IN HCl removed 
 all of the remaining silver from the five 
 resins in 6 h with less than 10 pet of 
 the gold being eluted. A final elution 
 with 0.75 pet NaClO in 150 g/L NaCl plus 
 5 g/L NaOH removed 93 to 98 pet of the 
 gold in 9 h. 
 
 Pregnant 
 
 leoch 
 
 solution 
 
 (Au, Ag, Hg) 
 
 Loaded SBR 
 
 Resin 
 
 adsorption 
 
 circuit 
 
 Barren 
 
 leach 
 
 solution 
 
 H2SO4- 
 
 NaCI plus HCl 
 
 NoCIO in 
 NoCI plus NoOh' 
 
 SBR 
 (Au,Ag,Hg) 
 
 SBR 
 {Au, Ag) 
 
 SBR 
 (Au) 
 
 ■Hg 
 
 -Ag 
 
 -Au 
 
 Barren SBR 
 
 FIGURE 1. - Flowsheet for recovery of mercury, 
 silver, and gold from a loaded strong-base exchange 
 resin (SBR) by sequential elution. 
 
TABLE 1. - Sequential elution of mercury, silver, and gold from strong-base 
 ion-exchange resins 
 
 
 Cumulative 
 elution 
 
 Metal eluted (cumulative) , pet 
 
 Eluant solution 
 
 Amberlite 
 
 Amberlite 
 
 Dowex 
 
 Dowex 
 
 Dowex 
 
 
 time, h 
 
 IRA-430 
 
 IRA- 9 00 
 
 21K 
 
 SMA-1 
 
 SBR 
 
 
 EG 
 
 Ag 
 
 Au 
 
 Hg 
 
 Ag 
 
 Au 
 
 Hg 
 
 Ag 
 
 Au 
 
 Hg 
 
 Ag 
 
 Au 
 
 Hg 
 
 Ag 
 
 Au 
 
 2N H2SO4 
 
 1 
 2 
 
 82 
 98 
 
 
 
 
 
 
 
 93 
 92 
 
 
 
 
 
 
 
 78 
 99 
 
 4 
 6 
 
 
 
 
 56 
 87 
 
 3 
 
 5 
 
 
 
 1 
 
 62 
 95 
 
 18 
 18 
 
 
 
 
 
 
 
 3 
 
 100 
 
 
 
 
 
 100 
 
 5 
 
 
 
 100 
 
 7 
 
 
 
 100 
 
 9 
 
 
 
 100 
 
 20 
 
 
 
 200 g/L NaCl in 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 IN HCl 
 
 4 
 5 
 
 100 
 100 
 
 62 
 85 
 
 
 2 
 
 100 
 100 
 
 65 
 89 
 
 
 
 
 100 
 100 
 
 69 
 88 
 
 
 
 
 100 
 100 
 
 57 
 84 
 
 2 
 2 
 
 100 
 100 
 
 62 
 83 
 
 n 
 
 
 
 
 
 6 
 
 100 
 
 95 
 
 5 
 
 100 
 
 97 
 
 
 
 100 
 
 96 
 
 1 
 
 100 
 
 93 
 
 2 
 
 100 
 
 93 
 
 
 
 
 7 
 
 100 
 
 98 
 
 7 
 
 100 
 
 99 
 
 
 
 100 
 
 98 
 
 2 
 
 100 
 
 96 
 
 5 
 
 100 
 
 97 
 
 3 
 
 
 8 
 
 100 
 
 100 
 
 8 
 
 100 
 
 100 
 
 
 
 100 
 
 100 
 
 4 
 
 100 
 
 98 
 
 5 
 
 100 
 
 99 
 
 3 
 
 
 9 
 
 100 
 
 100 
 
 8 
 
 100 
 
 100 
 
 
 
 100 
 
 100 
 
 4 
 
 100 
 
 99 
 
 5 
 
 100 
 
 100 
 
 3 
 
 0.75 pet NaClO in 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 150 g/L NaCl plus 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 g/L NaOH 
 
 10 
 
 100 
 
 100 
 
 75 
 
 100 
 
 100 
 
 83 
 
 100 
 
 100 
 
 31 
 
 100 
 
 100 
 
 28 
 
 100 
 
 100 
 
 68 
 
 
 11 
 
 100 
 
 100 
 
 83 
 
 100 
 
 100 
 
 91 
 
 100 
 
 100 
 
 58 
 
 100 
 
 100 
 
 56 
 
 100 
 
 100 
 
 79 
 
 
 12 
 
 100 
 
 100 
 
 87 
 
 100 
 
 100 
 
 95 
 
 100 
 
 100 
 
 79 
 
 100 
 
 100 
 
 67 
 
 100 
 
 100 
 
 87 
 
 
 13 
 
 100 
 
 100 
 
 87 
 
 100 
 
 100 
 
 95 
 
 100 
 
 100 
 
 91 
 
 100 
 
 100 
 
 76 
 
 100 
 
 100 
 
 90 
 
 
 14 
 
 100 
 
 100 
 
 88 
 
 100 
 
 100 
 
 97 
 
 100 
 
 100 
 
 93 
 
 100 
 
 100 
 
 85 
 
 100 
 
 100 
 
 91 
 
 
 15 
 
 100 
 
 100 
 
 89 
 
 100 
 
 100 
 
 97 
 
 100 
 
 100 
 
 95 
 
 100 
 
 100 
 
 92 
 
 100 
 
 100 
 
 92 
 
 
 16 
 
 100 
 
 100 
 
 91 
 
 100 
 
 100 
 
 97 
 
 100 
 
 100 
 
 96 
 
 100 
 
 100 
 
 98 
 
 100 
 
 100 
 
 94 
 
 
 17 
 
 100 
 
 100 
 
 92 
 
 100 
 
 100 
 
 97 
 
 100 
 
 100 
 
 96 
 
 100 
 
 100 
 
 98 
 
 100 
 
 100 
 
 94 
 
 
 18 
 
 100 
 
 100 
 
 93 
 
 100 
 
 100 
 
 98 
 
 100 
 
 100 
 
 97 
 
 100 
 
 100 
 
 98 
 
 100 
 
 100 
 
 95 
 
 WEAK-BASE RESINS 
 
 In addition to the work on sequential 
 elution of strong-base resins, research 
 was conducted with weak-base resins to 
 determine their effectiveness in adsorb- 
 ing gold and/or mercury from a caustic- 
 cyanide solution. Both commercially 
 available and experimental anion-exchange 
 resins were examined. Experimental 
 
 resins were included in the investigation 
 because they effectively adsorb metal- 
 cyanide complexes in the pH range in 
 which most gold mills operate, pH 10 to 
 11. Table 2 lists the weak-base resins 
 tested in this investigation, along with 
 available information concerning the res- 
 in matrix material and functional group. 
 
 TABLE 2. - Weak-base ion-exchange resins used in adsorption-elution 
 experiments, by manufacturer 
 
 Resin 
 
 Sybron Corp. : A-305 
 
 Diamond Shamrock Chemical Co: 
 
 A-7 , 
 
 A-340 , 
 
 A-561 , 
 
 Dow Chemical Co, : 
 
 MWA-1 , 
 
 WGR-2 , 
 
 XFS-40114 , 
 
 XFS-43309 
 
 XFS-43356 , 
 
 XU-40138 
 
 XU-40139 
 
 Functional group 
 
 Poly amine, 
 
 Secondary amine, 
 
 Polyamine , 
 
 ...do , 
 
 Dimethyl tertiary amine. 
 
 Polyamine 
 
 ...do , 
 
 NA. 
 NA. 
 NA. 
 NA. 
 
 Matrix 
 
 Epoxy amine. 
 
 Phenof ormaldehyde. 
 Epoxy amine, 
 Phenof ormaldehyde. 
 
 Styrene-DVB. 
 Epoxy amine. 
 Do. 
 
 NA. 
 NA. 
 NA. 
 NA. 
 
 NA Proprietary data not available from manufacturer. 
 
Batch-contact equilibrivim experiments 
 were performed with each Ion-exchange 
 resin to determine the effect of pH on 
 the resin adsorption of gold and/or mer- 
 cury. Six 1-g samples of each resin were 
 equilibrated at a specific pH value be- 
 tween 5 and 13 with NaOH and HCl. A syn- 
 thetic feed solution (200 mL) contain- 
 ing 500 ppm Au, 500 ppm Hg, and 0.5 g/L 
 NaCN was placed in a 250-mL plastic di- 
 gestion vessel, and the solution pH was 
 adjusted to the respective pH of each 
 resin. Resin was added and the mixture 
 was rolled for about 48 h. Following the 
 equilibrium contact, a sample of solution 
 was removed for gold and mercury analy- 
 sis, and the final equilibrium pH was 
 measured. 
 
 Figures 2 and 3 show the adsorption 
 results for gold and mercury, respective- 
 ly, for the commercially available res- 
 ins A-305, A-7, A-340, A-561, MWA-1, and 
 WGR-2 (as listed in table 2). The slopes 
 of these curves are very typical for 
 weak-base resins and show clearly the 
 influence of pH on a resin's ability to 
 adsorb metal-cyanide complexes. Better 
 adsorption occurred in the pH range below 
 9 for both gold and mercury, with none 
 of the resins selective for gold over 
 mercury. However, the A-305 resin indi- 
 cated a preference for loading mercury 
 over gold. 
 
 Additional research performed on the 
 A-305 resin indicated that this resin may 
 be suitable for application in adsorbing 
 mercury from existing gold operations. 
 Testing was expanded with a mill leach 
 solution to determine the effect of 
 adsorption-elution recycling using a 
 flow-through column and an experimental 
 procedure. Figure 4 presents typical ad- 
 sorption curves for both gold and mercury 
 using this resin. Initially, gold ad- 
 sorbed onto the resin, then quickly de- 
 sorbed into the solution. Mercury, how- 
 ever, steadily adsorbed. During the 
 recycling experiments, the data showed 
 that the resin's effectiveness to adsorb 
 mercury decreased with each successive 
 loading cycle. Attempts to regenerate 
 the resin were unsuccessful. 
 
 Subsequently, several experimental res- 
 ins were tested using the batch-contact 
 equilibrium technique previously de- 
 scribed. Figures 5 and 6 show the ad- 
 sorption results for gold and mercury, 
 respectively, for a group of experimental 
 resins developed by Dow Chemical Co. 
 (XFS-40114, XFS-43309, XFS-43356, XU- 
 40138, and XU-40139). Of particular in- 
 terest are the three resins XFS-40114, 
 XU-40138, and XU-40139. Gold and mercury 
 were more efficiently adsorbed by these 
 three resins, at a pH between 9 and 11, 
 than they were by any of the other previ- 
 ously examined resins. Also, the slopes 
 of the curves for these resins were much 
 steeper between pH 11 and 12, indicating 
 that the resins should be easily eluted 
 by a solution with a high pH value. Fol- 
 lowing the adsorption tests, a sample of 
 XFS-40114 was eluted using IM NaOH at am- 
 bient temperature. Nearly 100 pet of the 
 gold and 50 pet of the adsorbed mercury 
 were recovered with 100 mL of eluant. 
 
 Testing was expanded to determine the 
 effectiveness of the three promising res- 
 ins with actual leach solutions in a 
 flow-through contact column. The leach 
 solution contained 1.7 ppm Au, 3.0 ppm 
 Ag, and 0.3 ppm Hg, with a pH of approxi- 
 mately 10.5 and a free cyanide concentra- 
 tion of 0.5 g/L. Approximately 5 L of 
 solution was pumped through 1 g of resin 
 in a 1-cm-ID glass column at a flow rate 
 of 0.5 bed volume per minute. Following 
 the adsorption phase, the loaded resin 
 was eluted with 200 mL of IM NaOH at a 
 flow rate of 0.17 bed volume per minute. 
 Adsorption curves for gold, silver, and 
 mercury are shown in figures 7, 8, and 9, 
 respectively. The total adsorption and 
 elution recoveries for these experiments 
 are given in table 3. 
 
 TABLE 3. - Summary of adsorption-elution 
 results for selected resins, pet 
 
 Resin 
 
 Metal adsorption 
 
 Metal eluted 
 
 
 Au 
 
 Hg 
 
 Ag 
 
 Au 
 
 Hg 
 
 Ag 
 
 XFS-40114 
 
 XU-40138 
 
 XU-40139 
 
 54.2 
 57.7 
 81.1 
 
 51.8 
 57.4 
 66.6 
 
 4.8 
 10.8 
 13.2 
 
 13.7 
 53.1 
 13.1 
 
 
 
 8.7 
 
 6.0 
 
 76.6 
 85.1 
 51.9 
 
8 9 10 
 
 EQUILIBRIUM pH 
 
 FIGURE 2. - Equilibrium adsorption curves for gold with commercially available exchange resins. 
 (Broken curve patterns used for visual distinction only.) 
 
 8 9 10 
 
 EQUILIBRIUM pH 
 
 FIGURE 3..- Equilibrium adsorption curves for mercury with commercially available exchange resins. 
 (Broken curve patterns used for visual distinction only.) 
 
a. 
 
 CL 
 
 a. 
 
 Q- 
 
 <] 
 
 Q. 
 CC 
 O 
 (f) 
 Q 
 < 
 
 _l 
 UJ 
 
 0.4 0.8 1.2 
 
 SOLUTION VOLUME, L 
 
 FIGURE 4. - Gold and mercury adsorption curves 
 using resin A-305 with actual leach solution. 
 
 100 
 
 80 
 
 2 60 
 
 I- 
 
 Q. 
 CC 
 O 
 if) 
 
 § 40 
 
 o 
 
 20 - 
 
 
 
 1 'C::::;^-^..^^ ' 
 
 1 1 
 
 KEY 
 
 °\ ^^^^Sv\ 
 
 XFS-43356 
 
 \ ^nN/\ 
 
 □ XFS-43309 
 
 - h^ n\/\ 
 
 A XFS-401 14 - 
 
 \ ^vv ^ 
 
 . ■ XU-40138 
 
 \ ^k 
 
 \« XU-40 139 
 
 \ ^\ 
 
 
 \ 
 
 Ik^N. 
 
 \ 
 
 \\\ 
 
 \ 
 
 \ ^v\ 
 
 «v 
 
 V "^ 
 
 \ 
 
 \ ^\v 
 
 \ \ 
 
 \ IS. 
 
 s ^ 
 
 \ \\. 
 
 
 -a. \_AcN° 
 
 "^^ 
 
 \ \^^ 
 
 ~-~-o 
 
 ^-— dv ^ 
 
 1 1 1 
 
 1 1 
 
 7 
 
 12 
 
 8 9 10 II 
 
 EQUILIBRIUM pH 
 
 FIGURE 5. - Equilibrium adsorption curves for 
 gold with experimental exchange resins. 
 
 13 
 
 100 
 
 80 
 
 60 - 
 
 40 - 
 
 20 
 
 
 
 . ' "^ 
 
 1 
 
 1 
 
 
 \ 
 
 t^ 
 
 \\ 
 
 
 
 
 \ 
 
 \X 
 
 
 
 
 - ^\ 
 
 \?v 
 
 
 
 — 
 
 \\ 
 
 
 \ 
 
 
 
 \\ 
 
 V 
 
 \\ 
 
 
 
 - \ 
 
 1 
 
 \\\ 
 
 
 
 
 w ^ 
 
 \ 
 
 
 \\ 
 
 
 \\ 
 
 \ 
 
 
 \\ 
 
 
 v\ 
 
 
 \. 
 
 KEY \ 
 o XFS-43356 \> 
 
 
 \ 
 
 
 P 
 
 ■i 
 
 N^ 
 
 ^ 
 
 y 
 
 Q XFS-43309 
 
 ^^. 
 
 oS 
 
 ^ 
 
 
 -A XFS-401 1 4 
 ■ XU-40138 
 
 D 
 
 -=rtr- 
 
 \ 
 
 ^^^ 
 
 • XU-40139 
 
 1 1 
 
 1 
 
 1 
 
 
 1 
 
 8 
 
 9 10 II 12 
 
 EQUILIBRIUM pH 
 
 FIGURE 6. - Equilibrium adsorption curves for 
 mercury with experimental exchange resins. 
 
 13 
 
 CL 
 CL 
 
 CL 
 CL 
 <] 
 
 o 
 
 I- 
 
 Q. 
 
 q: 
 o 
 to 
 
 Q 
 < 
 
 O 
 
 uu 
 
 ^ 
 
 f 
 
 ■ "h— ^^-Ili* 1 
 
 ' 
 
 
 1 
 
 80 
 
 \ 
 
 
 • 
 
 \ 
 
 — 
 
 60 
 
 - 
 
 ^^ 
 
 =^^- 
 
 
 ^v • 
 
 
 
 KEY 
 
 ^^^ 
 
 ^^ 
 
 • N. 
 
 
 
 A XFS-401 14 
 
 
 "^ 
 
 IT"^^ ■ 1 
 
 40 
 on 
 
 
 ■ XU-40138 
 • XU-40139 
 
 1 1 
 
 1 
 
 
 1 
 
 I 2 3 
 
 SOLUTION VOLUME, 
 
 L 
 
 FIGURE 7. - Gold adsorption curves for resins 
 XFS-40n4, XU-40138, and XU-40139 with actual 
 leach solution. 
 
o 
 
 Q. 
 
 CL 
 Q. 
 
 80 
 
 60 
 
 Q. 
 CL 
 
 < 40 
 
 a. 
 cr 
 o 
 
 CO 
 Q 
 
 < 
 q: 
 
 LU 
 CO 
 
 20 
 
 
 
 KEY 
 A XFS-40 1 14 
 ■ XU-40138 
 • XU-40139 
 
 00 
 
 12 3 4 
 
 SOLUTION VOLUME, L 
 
 FIGURE 8. - Silver adsorption curves for resins 
 XFS-401 14, XU-40138, and XU-40139 with actual 
 leach solution. 
 
 Q- 
 
 
 Ci. 
 
 80 
 
 ^ 
 
 
 a 
 
 
 Q. 
 
 
 < 
 
 
 o 
 
 60 
 
 1- 
 
 
 LL 
 
 
 cr 
 
 
 o 
 
 
 CO 
 
 
 < 
 
 40 
 
 >- 
 
 
 en 
 
 
 3 
 
 
 o 
 
 
 cr 
 
 LU 
 
 20 
 
 Vn. 
 
 1 1 
 
 1 1 
 
 KEY 
 
 ^v^ 
 
 
 A XFS-401 14 
 
 -> 
 
 ^^ 
 
 ■ XU-40138 
 • XU-40139 
 
 — 
 
 
 K^ •\« • 
 
 ^^^^ 
 
 
 1 1 
 
 ■ 
 1 1 
 
 12 3 4 
 
 SOLUTION VOLUME, L 
 
 FIGURE 9. - Mercury adsorption curves for resins 
 XFS-40114, XU-40138, and XU-40139 with actual 
 leach solution. 
 
 These three resins (XFS-40114, XU- 
 40138, and XU-40139) were generally se- 
 lective for gold and mercury, with mini- 
 mal silver and other base metals being 
 adsorbed. The stronger the attraction a 
 resin had for gold and mercury, the less 
 selective it was and the poorer the elu- 
 tion with NaOH. The exchange resin XU- 
 40138 appeared to show the best loading 
 and eluting characteristics. During this 
 
 experiment, XU-40138 loaded to about 140 
 tr oz/st Au and produced an eluant con- 
 taining an average of 12.4 ppm Au. By 
 comparison, XU-40139 and XFS-40114 loaded 
 134 and 97 tr oz/st Au, respectively, and 
 eluted average Au concentrations of 1.7 
 and 3.9 ppm, respectively. The weak-base 
 resin XU-40138 may have a commercial ap- 
 plication if cyclic loading-elution tests 
 are successful. 
 
 SUMMARY AND CONCLUSIONS 
 
 Strong-base ani on-exchange resins were 
 loaded with gold, silver, and mercury, 
 then sequentially eluted with various 
 eluants. Sequential elution of Amberlite 
 IRA-430, for example, eluted 100 pet of 
 the mercury with 2N H2SO4, 100 pet of the 
 silver with 200 g/L NaCl in IN HCl, and 
 all of the remaining gold with 0.75 pet 
 NaClO in 150 g/L NaCl plus 5 g/L NaOH. 
 Depending on the strong-base resin exam- 
 ined, this elution scheme allowed little 
 contamination between the mercury and the 
 precious metals. 
 
 In addition to strong-base resins, 
 weak-base resins were investigated for 
 possible applications in the caustic- 
 cyanide systems of the gold industry. 
 None of the commercial weak-base resins 
 tested was suitable for a caustic-cyanide 
 system; however, three experimental res- 
 ins have possible application because of 
 their effectiveness in the pH range of 
 10 to 12. When using a mill leach solu- 
 tion, these resins adsorbed the gold and 
 mercury with only minimal silver being 
 adsorbed. 
 
REFEEIENCES 
 
 1. Arizona Pay Dirt. Gold Fields De- 
 velops State of the Art Leaching System 
 at Ortiz. No. 516, 1982, pp. 41-42, 44, 
 46, 48. 
 
 2. Engineering and Mining Journal. 
 Alligator Ridge Uses Heap Leaching To 
 Produce Gold Bullion Bars. V. 182, No. 
 8, 1981, pp. 35, 37. 
 
 3. Grace, K. A. Exploration and De- 
 velopment in 1981. World Min. , v. 183, 
 No. 7, 1981, pp. 58-62. 
 
 4. Jackson, A. Jerrit Canyon Project. 
 Eng. and Min. J., v. 183, No. 7, 1982, 
 pp. 54-58. 
 
 5. Skillings, D. N. , Jr. Getty Mining 
 Co. Starting Up Its Mercury Gold Opera- 
 tion in Utah. Skillings' Min. Rev., v. 
 72, No. 17, 1983, pp. 4-9. 
 
 6. . Homes take Proceeding With 
 
 Its McLaughlin Gold Project. Skillings' 
 Min. Rev., v. 72, No. 4, 1983, pp. 3-6. 
 
 7. . Pinson Mining Co. Mark- 
 ing First Full Year of Gold Production. 
 Skillings' Min. Rev., v. 71, No. 28, 
 1982, pp. 8-12. 
 
 8. . Smokey Valley Operations at 
 
 Its Round Mountain Mine in Nevada. 
 Skillings' Min. Rev., v. 68, No. 9, 1979, 
 p. 8. 
 
 9. Steel, G. L. Candelaria: Famous 
 Silver Producer. Min. Eng. (Littleton, 
 CO), V. 33, No. 6, 1981, pp. 659-660. 
 
 10. Laxen, P. A., G. S. M. Becker, and 
 R. Rubin. Developments in the Applica- 
 tion of Carbon-in-Pulp to Recovery of 
 Gold From South African Ores. J. S. Afr. 
 Inst. Min. & Metall. , v. 79, No. 11, 
 1979, pp. 315-326. 
 
 11. Potter, G. M., and H. B. Salis- 
 bury. Innovations in Gold Metallurgy. 
 (Pres. at Am. Min. Congr. Min. Conv. and 
 Environ. Show, Denver, CO, Sept. 9-12, 
 1973.) BuMines preprint (Salt Lake City, 
 UT), 1973, 12 pp. 
 
 12. Hussey, S. J. Application of Ion 
 Exchange Resins in the Cyanidation of a 
 Gold and Silver Ore. BuMines RI 4374, 
 1949, 34 pp. 
 
 13. Burstall, F. H. , P. J. Forrest, 
 N. F. Kember, and R. A, Wells. Ion Ex- 
 change Process for Recovery of Gold From 
 Cyanide Solution. Ind. and Eng. Chem. , 
 V. 45, No. 8, 1953, pp. 1648-1658. 
 
 14. Mooiman, M. D. , J. 0. Miller, 
 J. B. Hiskey, and A. R. Hendriksz. Com- 
 parison of Process Alternatives for Gold 
 Recovery From Cyanide Leach Solutions. 
 Paper in Proc. Soc. Min. Eng. AIME Fall 
 Meeting (Salt Lake City, UT, Oct. 19-21, 
 1983). AIME, 1984, pp. 93-107. 
 
 15. Von Michaelis, H. Innovation in 
 Gold and Silver Recovery. Soc. Min. Eng. 
 AIME, preprint 83-119, 1983, 9 pp. 
 
10 
 
 STAGED HEAP LEACHING-DIRECT ELECTROWINNING 
 By C. H. Elgesi and M. D. Wroblewski2 
 
 ABSTRACT 
 
 The Bureau of Mines is conducting re- 
 search to develop a staged, agglomeration 
 heap leaching system employing direct 
 electrowinning of dissolved precious met- 
 als from the pregnant leaching solution. 
 This paper describes bench-scale research 
 the Bureau conducted to develop such a 
 system and an efficient electrowinning 
 cell. 
 
 The Bureau demonstrated through leach- 
 ing simulations that pregnant solutions 
 
 suitable for direct electrowinning can be 
 generated by staged heap leaching. The 
 Bureau also demonstrated that improved 
 mass transfer (IMT) cells can recover 
 precious metals from low-grade pregnant 
 solutions generated by cyanide heap 
 leaching. This research indicates that 
 staged heap leaching-direct electrowin- 
 ning is a promising technique for recov- 
 ering precious metals from low-grade 
 resources. 
 
 INTRODUCTION 
 
 Heap leaching has been used to recover 
 gold and silver from low-grade ores since 
 the early 1970' s, and numerous operations 
 in the Western United States are present- 
 ly using heap leaching to produce pre- 
 cious metals (J:.~3.)»'^ One stimulus for 
 the rapid increase in the number of com- 
 mercial operations has been the dramatic 
 and sustained increase in the price of 
 gold. Another stimulus has been a suc- 
 cession of improvements in the practice 
 of heap leaching that has made possible 
 the treatment of increasingly difficult 
 ores. 
 
 Current heap leaching practice for pre- 
 cious metals includes sprinkling di- 
 lute alkaline NaCN solution on heaps of 
 crushed material, which may have been 
 agglomerated with cement or lime (4-^) , 
 and allowing the solution to percolate 
 through the heap. The effluent is col- 
 lected and passed through activated car- 
 bon beds that adsorb the precious metals. 
 For ores in which the silver content 
 is high relative to the gold content, 
 Merrill-Crowe zinc precipitation may be 
 
 ^Chemical engineer. 
 
 ^Physical science technician. 
 Reno Research Center, Bureau of Mines, 
 Reno, NV. 
 
 ^Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
 favorable ; this technique avoids the 
 large carbon inventories required to ad- 
 sorb the larger quantity of silver. When 
 carbon is used, the precious metals are 
 removed by stripping at elevated tempera- 
 ture with a strong caustic-cyanide solu- 
 tion, often with the addition of methyl 
 or ethyl alcohol (7). Gold and silver 
 are recovered from the stripping solution 
 by electrowinning onto a steel wool cath- 
 ode and are then fire-refined to produce 
 dore metal. 
 
 The carbon adsorption-desorption step 
 is an efficient way to concentrate the 
 typically dilute gold solutions (about 
 1 mg/L) from heap leaching. However, be- 
 cause the carbon is usually acid-washed 
 and must be thermally regenerated before 
 reuse, the carbon adsorption-desorption 
 step can add significantly to the operat- 
 ing and capital costs of a heap leaching 
 operation. In marginal cases, this step 
 may make an operation uneconomical. In 
 addition, there are precious metals 
 losses associated with loss of carbon 
 from the system. For these reasons, it 
 is desirable to eliminate the carbon 
 adsorption-desorption step and directly 
 electrowin the gold from the heap 
 effluent. 
 
 For staged heap leaching-direct elec- 
 trowinning to be efficient, the precious 
 metals concentration of the pregnant so- 
 lution generated during leaching must be 
 
11 
 
 higher than that obtained by conventional 
 heap leaching, and the solutions may have 
 to be fortified with electrolyte. The 
 precious metals concentration of the 
 pregnant solution can be increased by 
 (1) leaching with smaller volumes of so- 
 lution, (2) agglomeration pretreatment 
 with cyanide solution (which starts the 
 leaching process during the curing peri- 
 od) , and (3) cycling the leaching solu- 
 tion through more than one heap in a 
 staged manner. The resultant pregnant 
 solution is passed directly through an 
 electrowinning circuit. The barren solu- 
 tion from electrowinning is recycled to 
 the leaching circuit. 
 
 For direct electrowinning of precious 
 metals to be commercially feasible, it 
 is necessary to obtain suitable recov- 
 eries in electrolytic cells of reason- 
 able size and at ambient temperature, be- 
 cause heating large volumes of solution 
 is costly. This objective is made more 
 
 difficult because the concentration of 
 precious metals in heap leaching solu- 
 tions is very low, A typical heap leach- 
 ing pregnant solution contains 0,5 to 2.0 
 mg/L Au, compared to 50 to 1,000 mg/L Au 
 in carbon stripping solutions from which 
 gold is conventionally electrowon. The 
 requirements for direct electrowinning 
 are (1) precious metals electrowinning 
 cells that operate more efficiently at 
 ambient temperature than those currently 
 in use and (2) a method for conducting 
 heap leaching operations that maximizes 
 the precious metals content. The Bureau 
 is currently investigating direct elec- 
 trowinning of gold and silver from heap 
 leaching solutions and has developed 
 electrolytic cells with IMT characteris- 
 tics. Countercurrent staged heap leach- 
 ing techniques are being improved in 
 order to increase the concentration of 
 precious metals in the heap effluent. 
 
 DIRECT ELECTROWINNING 
 
 CELL DESIGN 
 
 The Bureau's original IMT cell design 
 is similar to that of the Zadra cell, 
 which has been the industry standard for 
 gold electrowinning for many years (8^) . 
 The Zadra cell (fig. 1) features three 
 concentric circular containers. The in- 
 ner container, which serves as the cath- 
 ode compartment, is a perforated insula- 
 tor that contains a central feed tube, 
 current distributor, and steel wool, onto 
 which precious metals are deposited. The 
 
 Pregnant solution • 
 
 CROSS SECTION 
 
 FIGURE 1. - Zadra cell design 
 
 Barren solution 
 SIDE VIEW 
 
 anode, a circular stainless steel screen, 
 is outside the cathode in the second con- 
 tainer. Pregnant solution enters through 
 the central feed tube and flows upward 
 and outward through the steel wool, 
 Zadra-type cells have been used to elec- 
 trowin precious metals from stripping 
 solutions at temperatures of 70° to 
 85° C. However, several disadvantages of 
 the Zadra design are that solution flow 
 is unevenly distributed, "effective" 
 electrode spacing is excessive,^ and cell 
 volume is utilized inefficiently. 
 
 The circular IMT cell (fig. 2) was 
 designed to overcome the problems of the 
 Zadra cell by providing uniform solution 
 flow across and through the cathodic 
 steel wool. The solution distribution 
 tube is stainless steel and serves as a 
 second anode, which decreases the effec- 
 tive electrode spacing. To increase the 
 cell efficiency, provision was made for 
 rapid internal solution circulation. Al- 
 though the mean residence time is unaf- 
 fected by this internal circulation, the 
 
 ^The physical properties of the pervi- 
 ous steel wool cathode preclude exact 
 measurement of electrode spacing. 
 
12 
 
 Pregnant solution 
 
 Recirculation pump 
 
 >> 
 
 Anode 
 
 n> 
 
 ® 
 
 ffi] 
 
 f3S;i 
 
 
 
 © 
 
 .Cathode 
 
 h 
 
 L 
 
 ■Barren solution 
 
 Cathode compartment 
 (steel wool) 
 
 Anode compartment 
 
 FIGURE 2. - Design of large IMT cell. 
 
 highly turbulent flow results in a thin- 
 ner electrode boundary layer, decreases 
 concentration polarization, and increases 
 metal deposition rates up to 200 pet. 
 Internal circulation flow rates were 10 
 to 25 times the pregnant solution feed 
 rate. The circular cell was built in 
 several sizes, ranging from 55 to 888 
 cm^ (cathode volume). 
 
 The data presented in this paper were 
 generated using circular IMT cells, al- 
 though current research efforts are con- 
 centrated on a rectangular IMT cell. The 
 rectangular cell is similar in appearance 
 to "breadbasket" types currently in use. 
 It should scale up easily while retaining 
 the favorable operating characteristics 
 of the circular cell, 
 
 ELECTROWINNING TESTS 
 
 Direct electrowinning tests were made 
 using small (55 cm^ cathode volume) and 
 large (888 cm^) circular IMT cells. The 
 cells were operated at ambient tem- 
 perature (20* to 25° C) with voltage, 
 amperage, feed rate, internal circula- 
 tion rate, NaOH concentration, and pre- 
 cious metals concentration as variables. 
 Small-cell tests were performed by pump- 
 ing solution from a reservoir through the 
 cell numerous times , using various flow 
 rates and operating times. Large-cell 
 tests were made by passing feed solution 
 through the cell once, but with internal 
 circulation of the solution. 
 
 For comparison, direct electrowinning 
 tests were made using small (67 cm^ cath- 
 ode volume) and large (888 cm^) Zadra 
 
 0.05 0.10 0.15 
 
 NaOH CONG, M 
 
 0.20 
 
 FIGURE 3. - 
 concentration. 
 72 mL ^min.) 
 
 Recovery of gold with increasing NaOH 
 (Cell operated at 3 V and feed rate of 
 
 cells. The cells were operated at 3 V 
 and ambient temperature (20° to 25° C) , 
 Retention time in the cells was varied, 
 
 RESULTS AND DISCUSSION 
 
 Operating data and results are given 
 for the small circular IMT cell in table 
 1 and for the large circular IMT cell in 
 tables 2 and 3, The results show that 
 flow rate and NaOH concentration are im- 
 portant parameters. Pregnant solutions 
 from heap leaching are unsuitable as cell 
 electrolytes unless they are fortified 
 with electrolyte. Figure 3 shows the in- 
 crease in gold recovery with the addition 
 of NaOH to 0,05M (4 Ib/st of solution) in 
 the large IMT cell. 
 
 Although NaOH was the most effective 
 additive for fortifying the electrolyte, 
 other salts were investigated, Na2C03 
 functioned well, but was not as effective 
 on a molar basis. Two salts of strong 
 acids, Na2S0i+ and NaN03 , were also inves- 
 tigated but caused severe corrosion of 
 the stainless steel anodes. 
 
 The effects of the circulation flow 
 rate on gold and silver recovery in the 
 small IMT cell are shown in figure 4, 
 Recovery of both metals increased mark- 
 edly with increased flow rates up 
 to approximately 250 mL/min and then 
 decreased. 
 
 The effects of the internal circulation 
 flow and feed rates on gold and silver 
 
13 
 
 TABLE 1. - Performance of small IMT celll 
 
 Current , 
 
 Operating 
 time, h 
 
 Flow rate, 
 mL/min 
 
 NaOH cone, 
 M 
 
 Recovery, pet 
 
 Deposition i 
 
 rate, mg/min 
 
 A 
 
 Au 
 
 Ag 
 
 Au 
 
 Ag 
 
 0.07 
 
 1.0 
 
 30 
 
 0.10 
 
 54 
 
 73 
 
 0.3 
 
 0.6 
 
 .08 
 
 1.0 
 
 75 
 
 .10 
 
 74 
 
 81 
 
 .5 
 
 .7 
 
 .08 
 
 1.0 
 
 150 
 
 .10 
 
 90 
 
 96 
 
 .6 
 
 .8 
 
 .09 
 
 1.0 
 
 240 
 
 .10 
 
 92 
 
 100 
 
 .8 
 
 1.0 
 
 .09 
 
 1.0 
 
 420 
 
 .10 
 
 84 
 
 99 
 
 .6 
 
 .9 
 
 .09 
 
 1.5 
 
 250 
 
 .10 
 
 97 
 
 100 
 
 .5 
 
 .7 
 
 .03 
 
 2.0 
 
 250 
 
 .01 
 
 43 
 
 26 
 
 .2 
 
 .1 
 
 .03 
 
 2.0 
 
 250 
 
 .02 
 
 46 
 
 31 
 
 .2 
 
 .1 
 
 .06 
 
 2.0 
 
 250 
 
 .05 
 
 61 
 
 78 
 
 .2 
 
 .4 
 
 .08 
 
 2.0 
 
 250 
 
 .10 
 
 84 
 
 91 
 
 .3 
 
 .4 
 
 .09 
 
 2.0 
 
 250 
 
 .15 
 
 92 
 
 92 
 
 .3 
 
 .4 
 
 .12 
 
 2.0 
 
 250 
 
 .20 
 
 96 
 
 96 
 
 .3 
 
 .5 
 
 ^Cell operated at 3 V. Pregnant synthetic solutions contained approximately 40 
 mg/L Au and 50 mg/L Ag. Cathode packing density, 0.018 g/cm^ (1 g steel wool). 
 
 TABLE 2. - Performance of large IMT cell showing effects of internal 
 circulation flow rate and feed rate^ 
 
 Current, 
 
 Retention 
 
 Internal circulation 
 
 Feed rate, 
 
 Recovery, 
 
 Deposition rate. 
 
 A 
 
 time, min 
 
 flow rate, L/min 
 
 mL/min 
 
 pet 
 
 mg/min 
 
 
 Au 
 
 Ag 
 
 Au 
 
 Ag 
 
 0.6 
 
 8.6 
 
 
 
 250 
 
 43 
 
 56 
 
 4.5 
 
 7.0 
 
 .7 
 
 8.6 
 
 .5 
 
 250 
 
 60 
 
 70 
 
 6.3 
 
 8.8 
 
 .7 
 
 8.6 
 
 1.0 
 
 250 
 
 67 
 
 78 
 
 6.5 
 
 10.3 
 
 .8 
 
 8.6 
 
 1.5 
 
 250 
 
 73 
 
 85 
 
 6.8 
 
 10.2 
 
 .8 
 
 8.6 
 
 2.0 
 
 250 
 
 75 
 
 81 
 
 7.4 
 
 9.1 
 
 .8 
 
 21.5 
 
 2.0 
 
 100 
 
 86 
 
 91 
 
 3.7 
 
 6.4 
 
 .8 
 
 4.3 
 
 2.0 
 
 500 
 
 51 
 
 63 
 
 8.1 
 
 16.0 
 
 ^Cell operated at 3 V with O.IM NaOH. Pregnant synthetic solutions contained 
 approximately 40 mg/L Au and 50 mg/L Ag. Cathode packing density, 0.018 g/cm^ (16 
 g steel wool) . 
 
 TABLE 3. - Performance of large IMT cell showing effects of cell voltage 
 and NaOH concentration^ 
 
 Potential, 
 
 Current, 
 
 NaOH cone. 
 
 PH 
 
 Au recov- 
 
 Deposition 
 
 Current efficiency. 
 
 V 
 
 A ■ 
 
 M 
 
 
 ery, pet 
 
 rate, mg/min 
 
 pet 
 
 1.5 
 
 ~o 
 
 0.05 
 
 12.75 
 
 
 
 
 
 
 
 2.0 
 
 .1 
 
 .05 
 
 12.75 
 
 39 
 
 1.8 
 
 23.4 
 
 2.5 
 
 .3 
 
 .05 
 
 12.75 
 
 64 
 
 2.0 
 
 12.9 
 
 3.0 
 
 .4 
 
 .05 
 
 12.75 
 
 90 
 
 2.8 
 
 14.8 
 
 3.5 
 
 .6 
 
 .05 
 
 12.75 
 
 95 
 
 2.9 
 
 9.8 
 
 4.0 
 
 .8 
 
 .05 
 
 12.75 
 
 97 
 
 3.0 
 
 7.6 
 
 3.0 
 
 .1 
 
 ~0 
 
 9.35 
 
 12 
 
 .4 
 
 7.3 
 
 3.0 
 
 .2 
 
 .01 
 
 10.45 
 
 68 
 
 2.1 
 
 21.0 
 
 3.0 
 
 .4 
 
 .05 
 
 12.75 
 
 90 
 
 2.8 
 
 14.8 
 
 3.0 
 
 .5 
 
 .10 
 
 13.05 
 
 94 
 
 2.9 
 
 11.5 
 
 3.0 
 
 .7 
 
 .15 
 
 13.13 
 
 96 
 
 3.0 
 
 8.5 
 
 3.0 
 
 1.0 
 
 .20 
 
 13.40 
 
 97 
 
 3.0 
 
 6.3 
 
 ^All tests used a retention time of 12.3 min, a feed rate of 72 mL/min, and an in- 
 ternal circulation flow rate of 2 L/min. Pregnant solutions contained 43 mg/L Au and 
 1 mg/L Ag. Cathode packing density, 0.018 g/cm^. 
 
14 
 
 100 
 ^ 90 
 
 / 
 
 /C^-^ 
 
 ^Silver 
 
 o 
 
 / 
 
 X 
 
 ^^^^^ _ 
 
 Q. 
 
 / / 
 
 ^r 
 
 ^© 
 
 .80 
 
 >- 
 
 '// 
 
 ^^Gold 
 
 
 uj 70 
 
 >/ 
 
 
 
 > 
 
 / 
 
 
 
 8 60 
 
 -/ 
 
 
 
 UJ 
 
 1 
 
 
 
 ^50 
 
 w 
 
 
 
 40 
 
 1 
 
 1 1 
 
 1 1 
 
 100 200 300 400 500 
 
 FLOW RATE, mL/min 
 
 FIGURE 4. - Recovery of gold and silver with in- 
 creasing circulation flow rate after 1 h of operating 
 time. (O.OlMNaOH.) 
 
 2 3 4 5 
 
 CELL POTENTIAL, V 
 
 FIGURE 5. - Recovery of gold with increasing cell 
 potential. (0.05M NaOH; feed rote, 72 mL/min; in- 
 ternal circulation flow rate, 2 L/min; retention time, 
 30 min.) 
 
 recovery in the large IMT cell are shovm 
 in table 2. Gold and silver recovery and 
 deposition rates increased as the inter- 
 nal circulation flow rate increased. 
 When the feed rate was increased, gold 
 and silver recovery decreased, but the 
 deposition rates increased. The data 
 show that silver is more easily electro- 
 won than gold; consequently, adjustment 
 of electrowinning parameters to achieve 
 the best gold recovery will ensure good 
 silver recovery. 
 
 The effects of decreasing gold con- 
 centration in the feed to the small 
 IMT cell are shown in table 4, Deposi- 
 tion rates decreased with decreasing 
 gold concentration, and the current effi- 
 ciency decreased to <1 pet when the gold 
 
 concentration in the pregnant solution 
 was less than 7 mg/L, 
 
 The effects of cell voltage on gold re- 
 covery and cell current in the large IMT 
 cell are shown in figure 5, Current var- 
 ied directly with applied voltage, and 
 gold recovery increased at greater cell 
 currents. Figure 6 shows that while gold 
 deposition rates increased with increased 
 cell potential, current efficiency de- 
 creased sharply. When the cell potential 
 was increased above 2 V, a greater pro- 
 portion of the current was used in the 
 decomposition of water as opposed to the 
 deposition of gold. 
 
 Table 5, which compares the opera- 
 tion of large Zadra and IMT cells, shows 
 that gold and silver recovery, current 
 
 TABLE 4. - Performance of small IMT cell with decreasing gold concentration 
 in feed solution^ 
 
 Au in 
 
 pregnant sol, 
 mg/L 
 
 Au recovered in 1 h of 
 operating time, pet 
 
 Deposition rate, 
 mg/min 
 
 Current efficiency, 
 pet 
 
 48 
 
 92 
 93 
 91 
 84 
 82 
 85 
 81 
 
 0.72 
 .60 
 .28 
 .11 
 .07 
 .03 
 .015 
 
 9.6 
 
 37 
 
 8.6 
 
 10 
 
 2.3 
 
 7.4 
 
 1.5 
 
 3.8 
 
 .71 
 
 2.1 
 
 .50 
 
 1.1 
 
 .20 
 
 ^Cell operated at 3 V, with a flow rate of 240 mL/min, a retention time of 3.3 min, 
 an NaOH concentration of O.IOM, and a cathode packing density of 0.018 g/cm^. 
 
TABLE 5. - Recovery of gold and silver in large electrolytic cells ^ 
 
 Feed rate, 
 mL/min 
 
 Current , 
 
 A 
 
 Synthetic 
 pregnant 
 solu t ion , mg 
 
 Au 
 
 Ag 
 
 Recovery, mg 
 
 Au 
 
 Ag 
 
 Current 
 
 efficiency, 
 
 pet 
 
 Au 
 
 Ag 
 
 Deposition 
 rate, 
 mg/min 
 
 Au 
 
 Ag 
 
 ZADRA CELL 
 
 100 
 
 0.60 
 
 592 
 
 720 
 
 290 
 
 403 
 
 2.5 
 
 6.3 
 
 1.8 
 
 2.5 
 
 250 
 
 .75 
 
 656 
 
 880 
 
 112 
 
 336 
 
 1.9 
 
 10.5 
 
 1.8 
 
 5.3 
 
 500 
 
 .75 
 
 640 
 
 896 
 
 64 
 
 208 
 
 2.2 
 
 12.9 
 
 2.0 
 
 6.5 
 
 IMT CELL 
 
 100 
 
 0.95 
 
 688 
 
 1,136 
 
 594 
 
 1,030 
 
 3.2 
 
 10.1 
 
 3.7 
 
 6.4 
 
 250 
 
 .95 
 
 608 
 
 816 
 
 440 
 
 629 
 
 5.9 
 
 15.4 
 
 6.9 
 
 9.8 
 
 500 
 
 .97 
 
 512 
 
 816 
 
 261 
 
 512 
 
 6.9 
 
 24.6 
 
 8.2 
 
 16.0 
 
 •'^Both cells were operated at ambient temperature, at 3 V, with a cath- 
 ode packing density of 0.018 g/cm^ , a feed volume of 16 L, and O.IM NaOH. 
 The IMT cell had an internal circulation flow rate of 2 L/min. Recovery 
 based on 1 pass through cell. 
 
 15 
 
 efficiency, and deposition rate, were all 
 substantially higher for the IMT cell 
 
 than for the Zadra cell , 
 higher feed rates. 
 
 especially at 
 
 STAGED HEAP LEACHING 
 
 HEAP PREPARATION AND LEACHING CYCLE 
 
 Staged heap leaching was developed to 
 produce pregnant solutions with the high- 
 est possible gold concentration. Figure 
 7 is a schematic of staged heap leaching 
 operated in conjunction with direct elec- 
 trowinning. Using this scheme, several 
 heaps are leached countercurrently be- 
 fore the pregnant solution is routed to 
 
 0.20 
 
 llJ 
 
 $ 
 
 1- 
 
 Q> 
 
 < 
 
 0) 
 
 a: 
 
 (0 
 
 
 ».— 
 
 z 
 
 o 
 
 O E 
 
 
 m 
 
 H 
 
 C3> 
 
 (D 
 
 h- 
 
 O 
 
 K 
 
 CL 
 
 r 
 
 Q 
 
 
 
 O) 
 
 
 E 
 
 .05- 
 
 2 3 4 
 CELL POTENTIAL, V 
 
 FIGURE 6. - Effect of cell potential on deposition 
 rate and current efficiency. (O.OSM^ NaOH; feed rate, 
 72 mL/min; internal circulation flow rate, 2 L/min, 
 retention time 30 min.) 
 
 electrowinning. The ore in the heaps is 
 agglomerated with cement and a strong 
 NaCN solution. The leaching action of 
 NaCN starts during the curing period and 
 may be almost finished by the time the 
 heap is constructed. The dissolved Au 
 and Ag can be removed from the heap by 
 repeated washing with a small volume of 
 water or dilute NaCN solution. Leaching 
 solution is applied intermittently be- 
 cause a "pulsed" flooding cycle resulted 
 in higher precious metals extraction and 
 used less leachant. 
 
 Final wash 
 
 NaCN ■ 
 Cement • 
 
 Agglomeration 
 
 Next heap 
 to be leached 
 
 FIGURE 7. - Staged heap leaching-direct 
 
 electrowinning. 
 
16 
 
 After leaching of a heap is completed, 
 a thorough washing cycle is conducted to 
 recover additional values. The wash wa- 
 ter is fortified with NaCN and used to 
 agglomerate a new charge of ore for heap 
 leaching. 
 
 LABORATORY TESTS 
 
 To conduct staged heap leaching-direct 
 electrowinning experiments in the labora- 
 tory, 22.7-kg (50-lb) charges of ore were 
 agglomerated and percolation leached in 
 acrylic columns 5-ft high by 5.5 in ID. 
 Agglomeration was accomplished with a 
 disk pelletizer in which the 22.7-kg ore 
 charges were combined with 227 g portland 
 cement (20 Ib/st ore) and 2.5 L of 0.1 OM 
 NaCN-0.05M NaOH solution (9.8 and 4.0 lb/ 
 st solution, respectively). The agglom- 
 erated ore was placed in columns and aged 
 for at least 24 h. A small IMT cell con- 
 taining 1 g of steel wool and operated at 
 3 V and a flow rate of 250 mL/min was 
 used to recover gold and silver from 
 solution. 
 
 Each test series used five columns, 
 each containing a 22.7-kg charge of ag- 
 glomerated ore. As shown in figure 8, 
 the laboratory procedure was to bring 
 columns on-line in stages , such that 
 steady-state operation could be approx- 
 imated by the completion of the test 
 series. In stage 1, 1 L of 0.05M NaOH 
 (4 Ib/st solution) was used to start 
 the leaching process; after progressing 
 to stage 3, there were three 1-L batches 
 
 of leachant in the system. Electrowin- 
 ning was conducted after each cycle of 
 leaching, and the barren cell electrolyte 
 was recycled to the next stage of leach- 
 ing. Additions of NaOH, when required, 
 were made prior to electrowinning, while 
 solution volumes were adjusted prior to 
 each leaching step. Upon completion of 
 leaching of columns 1 and 2, these col- 
 umns were subjected to wash cycles using 
 2.5-L quantities of water. The wash wa- 
 ters were then fortified with NaCN and 
 NaOH and used to agglomerate the ore 
 charges to columns 4 and 5, respectively. 
 Objectives throughout the leaching- 
 electrowinning sequence were to obtain 
 maximum precious metals recovery while 
 maintaining the greatest possible pre- 
 cious metals tenor in the pregnant leach 
 solutions. Pregnant solution analyses 
 are ideally based upon the average of the 
 pregnant leaching solutions produced dur- 
 ing stage 5. Precious metals recoveries 
 are ideally based upon the tails analysis 
 of column 3 after completion of stage 5. 
 A total of 25 to 30 cycles of leaching- 
 electrowinning were employed per test 
 series. 
 
 RESULTS AND DISCUSSION 
 
 Results of staged heap leaching-direct 
 electrowinning tests on four ores of dif- 
 fering grade and mineralogy are given 
 in table 6. The gold concentration of 
 the pregnant solutions produced was a 
 function of the ore grade. For the ores 
 
 STAGE 1 
 
 STAGE 2 
 
 STAGE 3 
 
 STAGE 4 
 
 STAGE 5 
 
 
 
 1 
 
 
 
 L 
 
 Electrowinning 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 2 
 
 
 
 L 
 
 — 
 
 
 
 
 
 
 Electrowinning 
 
 Electrowinning 
 
 ri 
 
 Electrowinning 
 
 n 
 
 Electrowinning 
 
 FIGURE 8. - Laboratory-scale staged leaching of agglomerated ores. (Numbers identify leaching 
 columns.) 
 
17 
 
 TABLE 6. - Recovery of gold by staged heap 
 leaching-direct electrowinnlng 
 
 Ore grade ppm Au.. 
 
 Gold recovery from ore, pet: 
 
 Staged heap leaching 
 
 Conventional leaching 
 
 Average content of pregnant 
 
 solution mg/L. . 
 
 NaOH used kg/mt ore. . 
 
 Current efficiency pet . . 
 
 Gold recovery, electrowinnlng, 
 pet: 
 
 Batch (average) 
 
 Overall 
 
 Test 
 
 9.3 
 
 87 
 
 86 
 
 32.5 
 
 0.43 
 7.3 
 
 89 
 98 
 
 A. 8 
 
 91 
 
 99 
 
 26.7 
 
 0.49 
 2.0 
 
 90 
 99 
 
 2.1 
 
 83 
 
 83 
 
 12.3 
 
 0.51 
 1.5 
 
 91 
 99 
 
 1.1 
 
 73 
 
 74 
 4.4 
 
 0.41 
 0.7 
 
 90 
 99 
 
 tested, the pregnant solutions generated 
 in each ease were at least an order of 
 magnitude higher in precious metals val- 
 ues than would be typical of solutions 
 generated using conventional leaching 
 practice. The increases in precious met- 
 als tenor were achieved in three of the 
 four tests with no measurable sacrifice 
 in precious metals recovery. Overall re- 
 covery of precious metals from solution 
 by electrowinnlng was more than 98 pet in 
 all eases. 
 
 Impurity buildup was not a problem dur- 
 ing the staged heap leaching-direct elec- 
 trowinnlng tests. Impurities were close- 
 ly monitored during test 4, which was 
 conducted on gold ore containing 1.1 g/mt 
 Au (0.035 tr oz/st Au) . Impurities that 
 accumulated in the pregnant feed to the 
 IMT cell were silicon (4 ppm) , mercury 
 (10 ppm), calcium (20 ppm), copper (21 
 ppm), and zinc (200 ppm). Only the mer- 
 cury codeposited with the precious 
 metals. 
 
 CONCLUSIONS 
 
 The research demonstrated that IMT 
 cells can recover precious metals from 
 low-grade pregnant solutions generated by 
 cyanide heap leaching. Efficient opera- 
 tion at ambient temperature makes the IMT 
 cell suitable for direct electrowinnlng. 
 Major differences in comparison with the 
 standard Zadra cell are the rapid recir- 
 culation of solution within the cell, 
 better control of solution flow streams, 
 and ambient temperature operation. 
 
 Simulated staged heap leaching demon- 
 strated that suitable pregnant solutions 
 for direct electrowinnlng can be gener- 
 ated. The buildup of impurities in the 
 recycled leachant and/or electrolyte did 
 not affect the electrowinnlng step in the 
 number of cycles studied. Staged perco- 
 lation leaching-direct electrowinnlng 
 shows considerable promise as a technique 
 for recovering precious metals from low- 
 grade resources. 
 
18 
 
 REFERENCES 
 
 1. Chamberlain, P. C, and M. G. Po- 
 jar. Gold and Silver Leaching Practices 
 in the United States. BuMines IC 8969, 
 1984, 47 pp. 
 
 2. Eisele, J. A,, A. F. Colombo, and 
 G. E, McClelland. Recovery of Gold and 
 Silver From Ores by Hydrometallurgical 
 Processing, Paper in Precious Metals: 
 Mining, Extraction and Processing. 
 (Proc. Int. Symp. held at AIME Annu. 
 Meeting, Los Angeles, CA, Feb. 27-29, 
 1984). AIME, 1984, pp. 387-395. 
 
 3. McQuiston, F. W. , and R. S. Shoe- 
 maker. Gold and Silver Cyanidation Plant 
 Practice, v. 2. Metall. Soc. AIME, 1981, 
 263 pp. 
 
 4. Heinen, H. J., G. E. McClelland, 
 and R. E, Lindstrom. Enhancing Perco- 
 lation Rates in Heap Leaching of Gold- 
 Silver Ores. BuMines RI 8388, 1979, 
 20 pp. 
 
 5. McClelland, G. E., and J. A. 
 Eisele. Improvements in Heap Leaching 
 To Recover Silver and Gold From Low- 
 Grade Resources. BuMines RI 8612, 1982, 
 26 pp. 
 
 6. McClelland, G. E., D. L. Pool, 
 and J. A. Eisele. Agglomeration-Heap 
 Leaching Operations in the Precious Met- 
 als Industry. BuMines IC 8945, 1983, 
 16 pp. 
 
 7. Heinen, H. J., D. G. Peterson, and 
 R. E, Lindstrom. Processing Gold Ores 
 Using Heap Leach-Carbon Adsorption Meth- 
 ods. BuMines IC 8770, 1978, 21 pp. 
 
 8. Zadra, J. B., A. L. Engel, and 
 H. J. Heinen. Process for Recovering 
 Gold and Silver From Activated Carbon 
 by Leaching and Electrolysis. BuMines 
 RI 4843, 1952, 32 pp. 
 
19 
 
 MERCURY PRECIPITATION DURING CYANIDE LEACHING OF GOLD ORES 
 
 By Richard G. Sandberg'' 
 
 ABSTRACT 
 
 Many gold-bearing ores throughout the 
 Western United States contain small quan- 
 tities of mercury. During cyanidation, 
 10 to 40 pet of the mercury is extracted 
 along with the precious metals. The 
 presence of mercury decreases gold load- 
 ing and increases stripping time on acti- 
 vated carbon, complicates fire refining 
 of the gold cathodes, and creates a pos- 
 sible health hazard. In the investiga- 
 tion described in this paper, the Bureau 
 of Mines examined several methods for re- 
 moving mercury from gold-silver cyanide 
 leach slurries. 
 
 CaS addition to cyanide leach slurries 
 or to a laboratory ball mill containing 
 
 NaCN and lime reduced mercury dissolution 
 to < 0.5 pet. Mercury loading on acti- 
 vated carbon was reduced to < 0.2 pet. 
 Gold loading on activated carbon was af- 
 fected very little by sulfide addition; 
 however, silver loading was reduced to 
 to 6 pet, as opposed to the typical val- 
 ues of 90 to 100 pet of the silver and 
 none of the mercury being adsorbed on the 
 carbon. 
 
 Preliminary testing in a mill operation 
 using NaHS showed that mercury precipi- 
 tation was nearly complete at the point 
 of addition; however, as with Na2S, 30 
 to 50 pet of the precipitated HgS redis- 
 solved with time. 
 
 INTRODUCTION 
 
 Mercury contamination has been a prob- 
 lem in the recovery of gold and silver 
 from many western deposits, which may 
 contain as much as 20 ppm Hg. During the 
 cyanide leaching process, 10 to 40 pet of 
 the mercury is normally solubilized along 
 with gold and silver. The mercury must 
 be recovered or precipitated so that it 
 does not present a health hazard during 
 electrolysis, smelting of the cathodes, 
 and regeneration of activated carbon. 
 Some gold mill operations recover mer- 
 cury by retorting the cathodes prior to 
 smelting (J^-2)^ or autoclaving the ore to 
 extract minimal mercury (3). 
 
 In an effort to reduce mercury solu- 
 bilization, the Bureau of Mines conducted 
 bench-scale leaching tests with an ore 
 containing 0.08 tr oz/st Au, 0.06 tr oz/ 
 St Ag, and 17 ppm Hg ( 4_) . Leaching was 
 accomplished with cyanide in an air- 
 agitated Pachuca-type vessel. Mercury 
 
 ^ Group supervisor, Salt Lake City Re- 
 search Center, Bureau of Mines, Salt Lake 
 City, UT. 
 
 ''Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
 extraction was decreased from 40 to 10 
 pet by (1) decreasing NaCN concentration 
 from 20 to 0.34 lb per short ton of solu- 
 tion, (2) decreasing the pH from 11,5 to 
 11, and (3) increasing particle size from 
 minus 270 mesh to minus 10 plus 48 mesh. 
 Although mercury extraction was reduced, 
 it was not eliminated. 
 
 The remaining soluble mercury can be 
 precipitated with sulfides as shown by 
 the chemical reaction 
 
 Hg(CN)^" + MS > HgS + m2+ + 4CN" , 
 
 Calcium, sodium, silver, iron, and zinc 
 sulfides have been used to precipitate 
 mercury from Au(CN)3 solutions (^~5^) • 
 Addition of Ag2S would tie up considera- 
 ble silver, and excess Ag2S would have to 
 be recovered. FeS is undesirable because 
 of cyanide loss due to ferrocyanide for- 
 mation. Calcium and sodium sulfides are 
 effective in precipitating mercury and 
 are not harmful to gold recovery. This 
 paper reports on a number of tests in 
 which calcium and sodium sulfides were 
 used to remove mercury from gold and 
 gold-silver cyanide leach slurries. 
 
20 
 
 MERCURY PRECIPITATION FROM SLURRY 
 
 Gold ore containing 17 ppm Hg was 
 leached with solutions containing 0.34 or 
 
 6 lb NaCN per short ton of solution and 
 enough lime to give a pH of 11, tjien con- 
 tacted with Na2S. The results of these 
 tests are shown in figure 1. Increasing 
 Na2S lowered mercury extraction when low 
 concentrations of cyanide (0.34 Ib/st 
 NaCN) were used. However, when high con- 
 centrations of cyanide (6 Ib/st NaCN) 
 were present, increasing the sulfide con- 
 centration in the solution increased mer- 
 cury extraction. This was probably due 
 to the formation of a soluble HgS2^" com- 
 plex (6^) . Because of this complex forma- 
 tion, high concentrations of Na2S have 
 been used to extract mercury from concen- 
 trates ij) , However, care must be taken 
 to control the amount of sulfide added, 
 since the addition of too much Na2S may 
 increase mercury extraction instead of 
 precipitating mercury as desired. 
 
 To determine the effects of Na2S con- 
 centration and time on mercury precipita- 
 tion (as HgS), additional tests were con- 
 ducted. Solutions containing 0.34 Ib/st 
 NaCN were contacted with gold ore for 24 
 h to extract 12 pet of the mercury; then 
 Na2S was added. Using ore containing on- 
 ly 0.02 Ib/st Na2S, 79.9 pet of the sol- 
 ubilized mercury was precipitated. How- 
 ever, within 0.5 h, the precipitated mer- 
 cury began to redissolve, and 4 h later, 
 nearly 30 pet of it had redissolved. 
 
 Because the HgS redissolved, CaS was 
 investigated as an alternative. In one 
 test, a cyanide solution containing 0.34 
 Ib/st NaCN (with no sulfides) extracted 
 12 pet of the mercury in 24 h. After 
 adding only 0.02 Ib/st CaS to this solu- 
 tion, 99.8 pet of the mercury was precip- 
 itated in only 0.5 h, and 7 h later, only 
 
 7 pet of the precipitated HgS had redis- 
 solved. A followup test, using 0.09 lb/ 
 St CaS, precipitated 100 pet of the solu- 
 ble mercury; after 24 h, < 0.01 pet of 
 the HgS had redissolved. 
 
 A comparison between HgS redissolution 
 using Na2S and CaS is shown in figure 2. 
 
 About 10 times more HgS redissolved with 
 Na2S than with CaS after 4 h. This may 
 be due to the formation of the soluble 
 Na2HgS2 complex (6^). CaS is less likely 
 to form this type of complex because of 
 its insolubility. 
 
 0.02 
 
 0.08 
 
 0.10 
 
 0.04 0.06 
 Na2S, Ib/st ore 
 
 FIGURE 1. - Effect of NojS and NaCN on mercury 
 extraction. 
 
 60 
 
 1 
 
 1 1 
 
 1 1 
 
 1 
 
 u 
 
 
 
 ^ 
 
 
 a. 
 
 
 ^^ 
 
 
 
 q" 
 
 
 ^y^ 
 
 
 
 UJ 
 
 
 /^NogS 
 
 
 
 rl 20 
 
 - 
 
 
 
 — 
 
 
 
 y^ 
 
 
 
 in 
 
 
 / 
 
 
 
 Q 
 
 
 
 
 
 UJ 
 
 
 
 
 
 cr 
 
 
 
 
 
 ^ 10 
 
 - X 
 
 
 
 — 
 
 z> 
 
 
 
 
 
 u 
 
 
 
 
 
 cc 
 
 LlI 
 
 
 
 ^_CaS_______ 
 
 — — " 
 
 2 
 
 ' \ ' 
 
 1 1 
 
 1 i 
 
 , 
 
 3 4 
 
 TIME, h 
 
 FIGURE 2. - Effect of time on sulfide precipita- 
 tion of mercury. 
 
21 
 
 MERCURY PRECIPITATION AND CARBON ADSORPTION 
 
 SIMULATED CARBON-IN-PULP 
 
 LABORATORY GRINDING CIRCUIT 
 
 Because most gold operations use acti- 
 vated carbon to recover precious metals, 
 tests were conducted to determine the ef- 
 fect of sulfide addition on carbon ad- 
 sorption of gold and mercury In a labora- 
 tory carbon-ln-pulp (CIP) circuit. Five 
 hundred grams of dry ground gold ore was 
 leached with 870 mL of leach solution 
 containing 0.34 ib/st NaCN and enough 
 lime to give a pH of 11. The slurry was 
 rolled for 24 h, then various amounts of 
 CaS were added and rolling was continued 
 for 1 h. Following HgS precipitation, 
 18 g/L activated carbon was added and 
 mixed for 1 h. The carbon was screened 
 from the slurry, and the slurry was con- 
 tacted with fresh carbon a second and 
 third time. 
 
 Results of these tests are listed in 
 table 1, Gold and silver adsorption on 
 carbon without prior CaS addition was 
 nearly 100 pet after the first stage, 
 while only 61 pet of the soluble mercury 
 adsorbed. Additional adsorption stages 
 recovered 9 pet more of the mercury. 
 With 0.023 ib/st CaS, mercury adsorption 
 was reduced to about 5 pet, without 
 decreasing gold and silver adsorption; 
 however, with 0.047 Ib/st CaS addition, 
 there was a slight decrease in silver ad- 
 sorption and mercury adsorption decreased 
 to 0.9 pet. 
 
 TABLE 1. - Carbon-ln-pulp adsorption 
 with CaS precipitation, percent' 
 
 Carbon stage 
 
 Au 
 
 Ag 
 
 Hg 
 
 Ib/st CaS: 
 
 1 
 
 99 
 1 
 
 
 
 99 
 1 
 
 
 
 99 
 .6 
 .1 
 
 97 
 3 
 
 
 96 
 
 4 
 
 
 95 
 4.1 
 .5 
 
 61 
 
 2 
 
 7 
 
 3 
 
 2 
 
 0.023 Ib/st CaS: 
 
 1 
 
 4.7 
 
 2 
 
 .4 
 
 3 
 
 .3 
 
 0.047 Ib/st CaS: 
 
 1 
 
 .37 
 
 2 
 
 .31 
 
 3 
 
 .20 
 
 Several gold operations add NaCN and 
 lime to the grinding circuit to extract 
 gold and silver. Because some of the 
 mercury is also extracted at the same 
 time, tests were conducted to determine 
 if addition of CaS to the ball mill would 
 prevent mercury extraction and how it 
 would affect gold, silver, and mercury 
 loading on carbon. The following were 
 added to a laboratory ball mill and 
 ground for 45 mln: 1,000 g of minus 10- 
 mesh ore, 0.5 Ib/st NaCN, 1,000 mL H2O, 
 enough lime to give a pH of 11, and vary- 
 ing amounts of CaS. Each resulting slur- 
 ry was washed into a 9-L bottle with 
 2,000 mL H2O and mixed. Gold and silver 
 were removed from the slurry by adsorp- 
 tion on activated carbon in three stages. 
 One gram of fresh carbon was used in each 
 stage. The total carbon contact time was 
 24 h. 
 
 The results of these tests are given in 
 table 2. Mercury extraction and adsorp- 
 tion on carbon were decreased from 15 to 
 0.4 pet and from 5.5 to 0.17 pet, respec- 
 tively, by increasing the amount of CaS 
 from 0.012 to 0.096 ib/st. Gold extrac- 
 tion decreased slightly, but gold loading 
 on carbon was essentially unaffected. 
 Silver extraction was unaffected, but 
 silver adsorption decreased. Both gold 
 and silver adsorption on carbon was lower 
 than in previous sulfide precipitation 
 tests. This was because much less car- 
 bon was used (0.33 g/L) than is normal- 
 ly used (18 g/L). Decreasing the carbon 
 
 TABLE 2. - Effect of CaS addition during 
 grinding on mercury, gold, and silver 
 extraction and adsorption, percent 
 
 
 Extraction' 
 
 Ad 
 
 sorption 
 
 CaS, Ib/st 
 
 Au 
 
 Ag 
 
 Hg 
 
 on carbon^ 
 
 
 Au 
 
 Ag 
 
 Hg 
 
 0.012... 
 
 99 
 
 86 
 
 15 
 
 91 
 
 71 
 
 5.5 
 
 .024 
 
 94 
 
 98 
 
 1.5 
 
 87 
 
 74 
 
 .55 
 
 .048 
 
 94 
 
 98 
 
 2 
 
 88 
 
 67 
 
 .40 
 
 .096 
 
 95 
 
 98 
 
 1.4 
 
 89 
 
 64 
 
 .17 
 
 'Percent of soluble metal (prior to CaS 
 addition) adsorbed on carbon. 
 
 'Total extraction including grinding 
 (45 mln) and carbon addition (24 h) . 
 ^24 h contact with carbon. 
 
22 
 
 concentration made it possible to analyze 
 for the small amount of mercury adsorbed 
 on the carbon. 
 
 GOLD- SILVER ORES 
 
 Because silver was precipitated dur- 
 ing mercury precipitation with CaS, addi- 
 tional tests were conducted. The pur- 
 pose of these tests was to determine the 
 effects of adding CaS and to find a way 
 to prevent silver precipitation. The 
 ores used in this study are listed in 
 table 3. 
 
 TABLE 3. - Analyses of gold-silver ores 
 
 Ore 
 
 Hg, 
 ppm 
 
 Ag, 
 tr oz/st 
 
 Au, 
 tr oz/st 
 
 Cu, 
 pet 
 
 Carline. . . . 
 
 Cortez 
 
 Silver Reef 
 
 16 
 9 
 3 
 
 2.52 
 .17 
 10.1 
 
 0.098 
 .013 
 .01 
 
 NAp 
 NAp 
 0.8 
 
 NAp Not applicable. 
 
 The ores were leached with 1 Ib/st NaCN 
 for 24 h; then CaS was added and mixed 
 in for 1 h. Following mercury precipi- 
 tation, the slurry was contacted with 
 activated carbon for 1 h. The test re- 
 sults are listed in table 4. Without 
 CaS addition, gold, silver, and mercury 
 adsorption values were all 100 pet; how- 
 ever, with CaS addition, both the silver 
 and mercury adsorption values were great- 
 ly reduced for all ores except the Silver 
 Reef ore. 
 
 Tests were conducted to determine why 
 silver was not precipitated from the 
 Silver Reef slurry during mercury pre- 
 cipitation. Close analysis of the Sil- 
 ver Reef pregnant leach solution showed 
 that it contained 170 ppm Cu. To elim- 
 inate silver loss during mercury pre- 
 cipitation, various amounts of CuCN 
 were added to a pregnant leach solution 
 containing 1,5 ppm Au, 3 ppm Ag, and 
 0.76 ppm Hg prior to the addition of 0.1 
 Ib/st CaS. The results given in figure 3 
 show that the addition of 160 ppm Cu 
 
 eliminated silver loss; but the copper 
 addition did not affect precipitation of 
 the mercury. 
 
 Additional tests were conducted as 
 described in table 4, except that 235 ppm 
 Cu was added prior to mercury precip- 
 itation. The results listed in table 5 
 show that silver recovery was greatly 
 increased. Carlin and Cortez silver 
 adsorption was increased from to 80 pet 
 and 6 to 91 pet, respectively. Mercury 
 adsorption was zero for all ores. 
 
 TABLE 4, - Adsorption on carbon in 
 simulated three-stage carbon-in-pulp 
 operation with CaS precipitation, 
 percent' 
 
 Carbon stage 
 
 Au Ag Hg 
 
 CARLIN ORE 
 
 Ib/st CaS: 
 
 1 
 
 90 
 9 
 
 1 
 
 50 
 
 41 
 
 9 
 
 91 
 9 
 
 
 
 
 
 
 100 
 
 2 
 
 
 
 3 
 
 
 
 0.10 Ib/st CaS: 
 
 1 
 
 8.1 
 
 2 
 
 
 
 3 
 
 
 
 CORTEZ ORE 
 
 Ib/st CaS: 
 
 1 
 
 57 
 30 
 13 
 
 64 
 25 
 11 
 
 75 
 
 25 
 
 
 
 6 
 
 
 
 100 
 
 2 
 
 
 
 3 
 
 
 
 0.10 Ib/st CaS: 
 
 1 
 
 2 
 
 2 
 
 
 
 3 
 
 
 
 SILVER REEF ORE 
 
 
 
 Ib/st CaS: 
 
 1 
 
 100 
 
 
 
 100 
 
 
 
 99 
 
 1 
 
 
 51 
 25 
 14 
 
 100 
 
 2 
 
 
 
 3 
 
 
 
 0.10 Ib/st CaS: 
 
 1 
 
 
 
 2 
 
 
 
 3 
 
 
 
 ' Percent of soluble metal adsorbed on 
 carbon. 
 
23 
 
 50 
 
 200 
 
 75 100 125 150 175 
 COPPER ADDITION, ppm 
 FIGURE 3. - Effect of copper addition on silver 
 precipitation during CaS precipitation of mercury. 
 
 TABLE 5. - Effect of copper addi- 
 tion on adsorption on carbon, 
 percent^ 
 
 (CaS addition: 0.1 Ib/st) 
 
 Carbon stage 
 
 Carlin ore 
 
 Au Ag Hg 
 
 Cortez ore 
 
 Au Ag Hg 
 
 WITHOUT COPPER ADDITION 
 
 1 
 
 50 
 
 41 
 
 9 
 
 
 
 
 
 8.1 
 
 
 
 
 
 64 
 25 
 11 
 
 6 
 
 
 
 
 ? 
 
 2 
 
 
 
 3 
 
 
 
 
 
 WITH 235 mg/L COPPER ADDITION 
 
 1 
 
 93 
 13 
 
 7 
 
 60 
 13 
 
 7 
 
 
 
 
 
 87.5 
 .8 
 .8 
 
 75 
 8 
 8 
 
 
 
 2 
 
 
 
 3 
 
 
 
 
 
 ^Percent of soluble metal adsorbed on 
 carbon. 
 
 MERCURY PRECIPITATION IN A MILL OPERATION 
 
 Precipitation of mercury with sulfides 
 is being tested at a northern Nevada mill 
 operation, using a process similar to 
 that illustrated in figure 4. Because 
 CaS was difficult to obtain, sodium hy- 
 drogen sulfide (NaHS) is being used in- 
 stead. A solution containing NaHS is 
 added to the NaCN slurry from the ball 
 mill before the slurry enters the thick- 
 ener. As shown in table 6 (feed to 
 thickener, days 3 to 6) , nearly all of 
 
 the mercury is precipitated at the point 
 of addition; however, over time, about 30 
 to 50 pet of this precipitated mercury 
 (HgS) redissolved. This was not unex- 
 pected, as NaHS acts similarly to Na2S, 
 and more NaHS addition points are re- 
 quired to obtain a more complete mercury 
 precipitation. Possible addition points 
 include the final leach tank prior to the 
 CIP tanks and the CIP tanks. 
 
 TABLE 6. - Mercury precipitation with NaHS in a mill operation 
 (Mercury analyses, parts per million^) 
 
 Sampling area 
 
 Sampling day 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 Feed to thickener 
 
 2.5 
 
 2,6 
 
 0.08 
 
 0,02 
 
 0.05 
 
 0.02 
 
 Feed to carbon column 
 
 
 
 
 
 
 
 (thickener overflow) 
 
 2.6 
 
 2.7 
 
 .09 
 
 .07 
 
 .11 
 
 .01 
 
 Final carbon column. ........ 
 
 2.6 
 
 2.8 
 
 .16 
 
 .03 
 
 .09 
 
 ,02 
 
 Feed to leach tanks (thick- 
 
 
 ener underflow) 
 
 2.5 
 
 2.8 
 
 ,68 
 
 .47 
 
 .83 
 
 ,19 
 
 Feed to carbon in pulp (from 
 
 
 
 
 
 
 
 leach tank 4) 
 
 2.7 
 
 2.9 
 
 3,3 
 
 1.3 
 
 1.8 
 
 1,3 
 
 Final carbon in pulp (to 
 
 
 
 
 
 
 
 tailings) 
 
 2,2 
 
 2.8 
 
 3,2 
 
 1.5 
 
 1.4 
 
 1,2 
 
 'Mercury remaining in solution. 
 
24 
 
 Ore »■ 
 
 NaHS 
 
 Carbon regeneration 
 kiln 
 
 Tailings 
 (HgS) 
 
 Return carbon 
 
 Mercury retort 
 
 Smelting furnace 
 
 Mercury 
 
 FIGURE 4. 
 operation. 
 
 Precious metals 
 - Flowsheet showing NaHS addition points for mercury precipitation in a mil 
 
CONCLUSIONS 
 
 25 
 
 CaS addition to gold-silver cyanide 
 leach slurries or to a laboratory grind- 
 ing circuit containing NaCN and lime re- 
 duced mercury dissolution to < 0,5 pet. 
 Mercury loading on activated carbon was 
 reduced to < 0.2 pet. Gold loading on 
 activated carbon was affected very little 
 by the sulfide addition; however, silver 
 loading appeared to be affected. Addi- 
 tional NaCN-CaS CIP tests with ores con- 
 taining greater amounts of silver result- 
 ed in only to 6 pet of the silver being 
 
 adsorbed by the carbon rather than the 
 normal 90 to 100 pet. Addition of copper 
 as CuCN, prior to the CaS addition, re- 
 sulted in 80 to 90 pet of the silver and 
 none of the mercury being adsorbed on the 
 carbon. 
 
 Preliminary testing in a mill operation 
 using NaHS showed mercury precipitation 
 was nearly complete at the point of addi- 
 tion; however, as had been found in other 
 tests with Na2S, 30 to 50 pet of the pre- 
 cipitated HgS redissolved with time. 
 
 REFERENCES 
 
 1. Craft, W. B. The Pinson Gold 
 Story. AZ Conf., AIME, Tucson, AZ, Dec. 
 6, 1982, 8 pp.; available upon request 
 from R. G. Sandberg, BuMines, Salt Lake 
 City, UT. 
 
 2. Skillings, D. N. , Jr. Getty Mining 
 Co. Starting Up Its Mercury Gold Opera- 
 tion in Utah. Skillings' Min. Rev., v. 
 72, No. 17, 1983, pp. 4-9. 
 
 3. . Homes take Proceeding With 
 
 Its McLaughlin Gold Project. Skillings' 
 Min. -Rev., v. 72, No. 4, 1983, pp. 3-6. 
 
 4. Sandberg, R. G. , W. W. Simpson, 
 and W. L. Staker. Calcium Sulfide 
 
 Precipitation of Mercury During Cyanide 
 Leaching of Gold Ores. BuMines RI 8907, 
 1984, 13 pp. 
 
 5. Flynn, C. M., Jr., T. G. Carnahan, 
 and R. E. Lindstrom. Selective Removal 
 of Mercury From Cyanide Solutions. U.S. 
 Pat. 4,256,707, Mar. 17, 1981. 
 
 6. Habashi, F. Hydrometallurgy . 
 Principles of Extractive Metallurgy, v. 
 2. Gordon and Breach, 1969, p. 99. 
 
 7. Town, J. W. , and W. A. Stickney. 
 Cost Estimates and Optimum Conditions for 
 Continuous-Circuit Leaching of Mercury, 
 BuMines RI 6459, 1964, 28 pp. 
 
26 
 
 CARBONACEOUS GOLD ORES'" 
 By B, J. Schelner2 
 
 ABSTRACT 
 
 The presence of organic material in 
 gold ores interferes with gold extrac- 
 tion by cyanidation. Oxidation of the 
 organic material using chlorine-hypo- 
 chlorite systems has been shown to render 
 the gold ore amendable to cyanidation. 
 This paper discusses three oxidation 
 procedures that have been investigated 
 by the Bureau of Mines: (1) addition of 
 
 sodium hypochlorite (NaOCl) to ore pulp, 
 
 (2) addition of chlorine to ore pulp, and 
 
 (3) in situ generation of NaOCl by elec- 
 trolysis of a brine solution used to pulp 
 the ore. Gold extraction was greater 
 than 90 pet when carbonaceous ore was 
 oxidized and then cyanided. Both labo- 
 ratory and pilot plant results are 
 discussed. 
 
 INTRODUCTION 
 
 The presence of carbon and organic com- 
 pounds that inhibit gold recovery from 
 auriferous ores has long plagued cyanide 
 mill operators. Research to develop 
 techniques for testing these ores, which 
 are commonly referred to as carbonaceous 
 ores, was conducted in the early 1920' s 
 by the Bureau of Mines (O.-^ Research 
 was continued by the Bureau and others, 
 but as late as 1958, it was indicated 
 that no solution to the carbonaceous 
 problem was at hand (2^). After the dis- 
 covery of carbonaceous gold ores at the 
 Carlin Mine in northeastern Nevada, the 
 Bureau initiated an extensive research 
 program in 1966 to solve the problems as- 
 sociated with extracting gold from these 
 ores. 
 
 Research was conducted by the U.S. 
 Geological Survey, the Bureau of Mines, 
 and Newmont Mining Corp., the owners of 
 the Carlin Gold Mine, to characterize 
 the carbonaceous ores found in Nevada 
 (3-5). Early studies indicated that the 
 
 gold-bearing sedimentary beds were asso- 
 ciated with the Roberts Mountain thrust 
 fault system, generally in windows expos- 
 ing lower plate sedimentary beds of Silu- 
 rian age. However, later discoveries 
 have shown that the ore association with 
 the Roberts Mountain thrust fault system 
 is just a fortuitous anomaly, since gold 
 has been found in upper plate material 
 outside the Roberts Mountain thrust 
 fault. It has been hypothesized that the 
 gold was redistributed and concentrated 
 in permeable carbonaceous horizons by 
 hydrothermal solutions. Apparently the 
 same solutions replaced much of the silty 
 dolomitic limestone with microcrystalline 
 quartz. Subsequent oxidation induced by 
 shallow meteoric oxygen-bearing waters 
 removed the carbonaceous material from 
 the upper portions of the deposits, thus 
 oxidizing these portions of the deposits. 
 The Carlin Gold Mine property is an exam- 
 ple of a deposit that contains both oxi- 
 dized and carbonaceous gold ores. 
 
 CYANIDATION OF CARBONACEOUS ORES 
 
 To determine the effect of carbonace- 
 ous ores on cyanidation, samples of both 
 oxidized and carbonaceous ore were ob- 
 tained from various locations in Nevada, 
 including the Carlin Gold Mine, Cortez 
 
 ^This paper is based upon work done un- 
 der an agreement between the University 
 of Alabama and the Bureau of Mines, 
 
 ^Supervisory metallurgist, Tuscaloosa 
 Research Center, Bureau of Mines, Univer- 
 sity, AL. 
 
 Gold Mine, and the abandoned Gold Acres 
 pit at Crescent Valley. The gold content 
 of the carbonaceous ores ranged from 0.23 
 to 0.40 tr oz/st Au, with an organic 
 carbon content ranging from 0.3 to 0,6 
 wt pet. 
 
 Initial experiments involving cyanida- 
 tion of the carbonaceous ores over a wide 
 
 ^Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
27 
 
 10 20 30 40 50 
 
 CARBONACEOUS ORE ADDED TO OXIDE ORE, wt pet 
 
 FIGURE 1. - Effect on gold extraction of adding 
 carbonaceous ore to oxide ore. 
 
 range of operating conditions showed that 
 only 5 to 32 pet of the gold was amenable 
 to recovery by this conventional tech- 
 nique. The data indicated that cyanide 
 was being consumed, but not excessively. 
 The oxygen content of the leach solution 
 was determined to be favorable for effec- 
 tive cyanidation. Analysis of washed 
 tails detected cyanide, suggesting that 
 Au(CN)2- was being adsorbed on carbonace- 
 ous components of the ore. The possibil- 
 ity of gold adsorption by carbon was fur- 
 ther investigated by contacting 0.1 tr/st 
 of pregnant Au(CN)2- solutions with vari- 
 ous carbonaceous ores. From 12.5 to 140 
 tr oz Au was adsorbed per short ton of 
 carbonaceous ore. 
 
 To investigate how the addition of 
 carbonaceous material affects the cya- 
 nidation of oxidized ores, a series of 
 experiments was conducted in which 
 various amounts of carbonaceous material 
 were added to oxidized ore and the sam- 
 ple cyanided. The results of these 
 experiments are shown in figure 1. Gold 
 extraction decreased linearly as the car- 
 bonaceous ore content increased (_6-_7) . 
 Application of the well-known technique 
 of blanking out an activated carbon com- 
 ponent by coating it with kerosene or 
 some other petroleum product was shown 
 to be only partially effective, depend- 
 ing on the particular carbonaceous ore 
 being treated. Treatment of the ore 
 with anion-exchange resins or granulated 
 
 activated carbon during cyanidation 
 in order to actively compete for the 
 Au(CN)2- complex generally improved the 
 gold extraction, but the results varied 
 considerably depending on the particular 
 sample being used. This suggested that 
 certain samples of the carbonaceous mate- 
 rial contained two different types of 
 carbon that prevented favorable extrac- 
 tion: (1) an activated type of carbon 
 that physically adsorbed the Au(CN)2- 
 complex, and (2) a hydrocarbon type that 
 formed a gold compound during deposition 
 that was not attacked by cyanide (^-7^) • 
 
 Gold extraction depended markedly on 
 the basicity of the system, indicating 
 the presence of carboxylic acid groups 
 that saponified in the highly basic ion- 
 exchange system. Isolation of organic 
 compounds in the sample was accomplished 
 by an aqueous NaOH treatment for 2 h, 
 followed by acidification of the liquor 
 and solvent extraction of the dissolved 
 organic compounds into chloroform (3) . 
 Infrared spectra of the organic extrac- 
 tion product had major adsorption peaks 
 at 2,900 and 1,700 cm~^ , which were char- 
 acteristic of long-chain carboxylic 
 acids. The neutralization equivalent of 
 this mixture of organic acids was 1,500. 
 Sulfur and nitrogen groups were also 
 found in the organic material. The 
 amounts of organic material that could 
 be accumulated from the sample were only 
 sufficient for identification purposes. 
 The extracted organic compounds were 
 found to be remarkably similar to humic 
 acid extracted from leonardite that oc- 
 curs in North Dakota (8^). These humic 
 acid extracts contain long-chain carbox- 
 ylic acids as well as sulfur and nitrogen 
 groups. Cyanidation of oxidized ore in 
 the presence of humic acids showed that 
 Au(CN)2- was adsorbed or associated with 
 the humus compounds and that gold recov- 
 ery was essentially nil. 
 
 It was concluded that the low extrac- 
 tion of gold from the carbonaceous mate- 
 rials could not be attributed wholly to 
 physical adsorption of the Au(CN)2- com- 
 plex on carbon, but that in certain of 
 these materials, a substantial amount of 
 the gold was locked in in the form of a 
 chelate containing CO-N-S ligands. 
 
28 
 
 CHEMICAL OXIDATION OF CARBONACEOUS ORES 
 
 LABORATORY EXPERIMENTS 
 
 The data indicated that destruction or 
 complete pacification of detrimental car- 
 bon components is a requisite for effec- 
 tively extracting gold from the various 
 carbonaceous materials. It was deter- 
 mined that a mild chemical oxidation of 
 the carbonaceous ores followed by cya- 
 nidation yielded high gold extraction. 
 Various oxidation systems were investi- 
 gated, including ozone, chlorine, NaOCl, 
 calcium hypochlorite [Ca(0Cl)2], perman- 
 ganates, perchlorates, chlorates, and 
 oxygen. Gold extraction, obtained by 
 cyanidation after oxidation, increased 
 significantly in nearly every case over 
 that obtained without pretreatment ( 6-_7 ) . 
 
 The use of ozone as an oxidant was fur- 
 ther investigated (9^). It was determined 
 that 95 pet Au extraction could be ob- 
 tained from several different carbonace- 
 ous ores by slurrying the ore with brine 
 solution, lowering the pH to 1 with H2S0i^ 
 or HNO3 , and bubbling ozone through the 
 slurry. After the ozone treatment was 
 completed, the gold was extracted by cya- 
 nidation. However, owing to the acid- 
 consuming calcareous nature of the ore, 
 the ozone system, in which an acid medium 
 was used, was ruled out as an impractical 
 solution to the carbonaceous problem. 
 
 Introduction of chlorine gas into an 
 ore slurry, followed by filtering and 
 subsequent cyanidation of the filter 
 cake, resulted in 95 pet Au extraction 
 from several carbonaceous ores (_7 ) . 
 Rapid addition of the chlorine gas (as 
 represented below) caused the pH of the 
 ore pulp to drop to the range of pH 1 
 to 2. 
 
 CI2 + H2O ->■ HCl + HOCl. 
 
 (A) 
 
 However, the pH remained neutral with 
 slow addition of chlorine gas because the 
 HCl formed reacted slowly with the cal- 
 careous gangue material to give CaCl2 : 
 
 2HC1 + CaC03 ->■ CaClg + H2CO3. (B) 
 
 The active oxidizing species obtained 
 from rapid addition of chlorine gas 
 is principally HCl. Slow addition of 
 
 chlorine gas results in production of 
 Ca(0Cl)2 as the principal oxidizing 
 species: 
 
 2H0C1 + CaC03 -»■ Ca(0Cl)2 + H2CO3. (C) 
 
 The amount of chlorine gas that escapes 
 from the pulp can be excessive. To over- 
 come this problem, NaOH or lime was added 
 to the slurry before the addition of 
 chlorine gas. The chlorine gas reacts 
 with the base to form Ca(0Cl)2 or NaOCl: 
 
 CI2 + 2NaOH -»- NaOCl + NaCl + H2O, (D) 
 
 2CI2 + 2Ca(0H)2 ->■ Ca(0Cl)2 
 
 + CaCl2 + 2H2O. (E) 
 
 The hypochlorite or oxidizing species 
 then reacts with carbonaceous material in 
 the ore, passivating activated-type car- 
 bon and breaking hvdrocarbon chains and 
 htunic acid-type components. 
 
 Since the NaOCl (household bleach) 
 proved to be an effective oxidizing 
 agent, the parameters affecting gold ex- 
 traction were investigated. It was de- 
 termined that 90 pet Au extraction could 
 be obtained at 50° C in 4 h using 16 to 
 20 lb NaOCl per short ton of ore along 
 with 20 Ib/st of lime (7^). 
 
 Instead of adding NaOCl or chlorine to 
 the ore pulp, the oxidizing condition 
 could be produced by slurrying the ore in 
 brine and electrolyzing to produce NaOCl 
 in situ. A literature survey indicated 
 that the technology for the electrolysis 
 of NaCl in ore slurries was essentially 
 nonexistent. Preliminary experiments in- 
 dicated that oxidizing conditions could 
 be generated and controlled under a vari- 
 ety of conditions by this electrooxida- 
 tion concept (10-11). Parameters deemed 
 to be important to the development of the 
 technique included salt concentration, 
 electrolysis time at constant amperage 
 per short ton of ore, temperature, cur- 
 rent density, type of electrodes, elec- 
 trode spacing, and particle size of ore 
 
 (I). 
 
 During the course of the investigation 
 it was determined that a variety of elec- 
 trode materials and configurations could 
 
29 
 
 be used. The only difficulty encountered 
 was a buildup of a deposit on the cathode 
 with time. This difficulty was overcome 
 by using graphite electrodes and revers- 
 ing the current periodically to remove 
 the deposits. One type of cell used, a 
 plate-type laboratory-scale electrooxida- 
 tion cell, is shown in figure 2. 
 
 The effect of salt concentration (NaCl) 
 on gold extraction from carbonaceous ore 
 was investigated. It was determined that 
 an 8- to 10-pct-salt concentration was 
 adequate for obtaining sufficient oxida- 
 tion to result in efficient gold extrac- 
 tion by cyanidation. 
 
 The effect of temperature on gold ex- 
 traction was determined through tests at 
 30°, 40°, and 50° C. Maximum gold ex- 
 traction was obtained at 40° C. Heat in- 
 put to an electrooxidation system and the 
 resulting temperature are functions of 
 the conductivity of the electrolyte and 
 the power required to accomplish the oxi- 
 dation; therefore, parameters such as 
 electrode spacing, salt concentration, 
 and pulp density are all critical in 
 maintaining the desired temperature. 
 Generally speaking, electrode spacing 
 should be as close as is consistent with 
 
 Busbar 
 
 Pulp level 
 
 Wood 
 support 
 
 Graphite 
 cathode 
 
 Graphite 
 anode 
 
 Ore slurry 
 
 30-50 pet 
 
 solids 
 
 good pulp flow between electrodes. The 
 effect of increasing electrode spacing 
 was investigated. Spacings of 3/8, 5/8, 
 and 1-1/8 in were used, with pulp den- 
 sity at 40 pet and a salt concentration 
 of 10 pet. As was expected, the resist- 
 ance between the electrodes increased 
 with increased electrode spacing. The 
 effect of pulp density on conductance was 
 also studied (fig. 3). As was expected, 
 the conductance decreased rapidly as pulp 
 density increased. 
 
 Current density is another factor 
 that affects the voltage-amperage 
 relationship, which in turn affects the 
 temperature. The effect of current den- 
 sity on the voltage required to maintain 
 constant amperage during electrolysis is 
 shown in figure 4. The voltage increased 
 linearly with current density over the 
 range measured. The data indicated that 
 the current density should be as low as 
 possible to keep power consumption at a 
 minimum. However, the current density 
 dictates the number of electrodes re- 
 quired, and the tonnage of the mill and 
 the size of the agitators employed are 
 important factors that must also be con- 
 sidered in projecting the number of elec- 
 trodes that can be practically utilized. 
 
 Grinding is an important part of any 
 hydrometallurgical process; it releases 
 
 0.18 
 
 FIGURE 2. - Plate-type laboratory-scale electro- 
 oxidation cell and agitation vessel. 
 
 20 30 40 50 60 
 
 PULP DENSITY, pot 
 
 FIGURE 3. - Effect of pulp density on conductance 
 in ore-brine pulp (using 9.77-pct-salt solution). 
 
30 
 
 0.4 0.8 1.2 1.6 2.0 
 
 ANODE CURRENT DENSITY, A/in^ 
 
 FIGURE 4. - Effect of anode current density on 
 voltage at constant amperage. 
 
 the mineral from the host rock so the 
 mineral can come in contact with the re- 
 actants. The effect of particle size on 
 gold extraction was investigated using 
 several different grinds based on the 
 percentage of minus 200-mesh material. 
 Data for one ore indicated that gold ex- 
 traction increases as the particle size 
 becomes smaller and that a grind of 70 
 pet minus 200 mesh was satisfactory; how- 
 ever, each ore must be evaluated on an 
 individual basis as to optimum particle 
 size for reaction. 
 
 PILOT PLANT STUDIES 
 
 Based on the results obtained in labo- 
 ratory studies, the use of NaOCl and 
 chlorine and the in situ generation of 
 NaOCl (electrooxidation) were investi- 
 gated on a pilot plant scale (12-13). 
 The ores used in these studies were ob- 
 tained from the Carlin Gold Mine. The 
 flow sequence and pilot plant operation 
 generally followed conventional counter- 
 current-decantation slime-circuit operat- 
 ing practice. Ore material was ground to 
 60 pet minus 200 mesh at 50 pet solids in 
 the rod mill and pulped into the oxida- 
 tion tanks, each of which held 275 lb of 
 dry ore (550 lb of pulp). After treat- 
 ment in the oxidation tank at the desired 
 temperature, the pulp was passed through 
 the digestion-surge tank, where 1 lb of 
 cyanide per short ton of ore was added, 
 along with sufficient lime to maintain a 
 cyanidation pH of 11. Approximately 5 lb 
 of lime per short ton of ore was usually 
 
 required to maintain the desired pH val- 
 ue. The pulp was then pumped through 
 three cyanidation tanks, for a cyanida- 
 tion time of 9 h. From the cyanidation 
 tanks the pulp passed through four thick- 
 eners at 20 pet solids, flowing counter- 
 current to the barren solution recycled 
 from the gold-precipitation sequence. 
 Pregnant solution was passed from the 
 first thickener to the gold-precipitation 
 system. 
 
 Initial pilot plant experiments were 
 conducted on nonrefractory oxide gold ore 
 to determine the best operating condi- 
 tions for the grinding circuit and pulp 
 handling, optimum flow rates, etc. Mill 
 tails from the oxide ore contained 0.008 
 tr oz/st Au, which corresponds to a gold 
 extraction of 96 pet. 
 
 Carbonaceous gold ore was processed in 
 the pilot mill, using conventional cyani- 
 dation, without oxidation pretreatment , 
 as a baseline experiment to determine the 
 effect of carbonaceous material. Gold 
 extraction obtained in these experiments 
 ranged from 29 to 33 pet (0.26 to 0.28 
 tr oz Au per short ton of tails). 
 
 Oxidation by NaOCl Addition 
 
 The pilot plant operations generally 
 followed the operating conditions estab- 
 lished in the laboratory experiments for 
 obtaining favorable gold extraction val- 
 ues. The oxidation section of the plant 
 was operated on a semicontinuous basis 
 with ground ore being treated on a batch 
 basis in successive oxidation tanks. The 
 ore then could be discharged into the 
 surge tank in such a manner that a re- 
 serve of treated pulp was available for 
 continuous operation of the cyanidation 
 and liquid-solids separation sections of 
 the mill. The oxidation section of the 
 mill processed 80 lb of dry ore per hour, 
 thus providing a retention time of 9 h in 
 the cyanidation tanks. 
 
 Figure 5 shows the oxidation results — 
 the gold extractions obtained at differ- 
 ent levels of NaOCl addition to the pulp 
 containing carbonaceous ore. As shown in 
 figure 5, the reaction time, temperature, 
 and lime addition were kept constant. 
 The pulp pH values were in the 11 to 11.5 
 range. Extraction increased markedly 
 
31 
 
 90 
 
 ' ' ^— — 
 
 - 
 
 _ / Conditions: 
 
 
 / Reaction time, 4 h 
 
 
 / Lime, 10 Ib/st 
 
 
 / Temp, 50° to 60° C 
 
 
 /ill 
 
 — 
 
 ^ 80 
 o 
 
 a. 
 
 o 70 
 
 I- 
 
 ^ 60 
 
 X 
 
 UJ 
 
 Q 50 
 
 _i 
 o 
 o 
 40 
 
 30. 
 
 5 10 15 20 
 
 NoOCI ADDITION, Ib/st 
 
 FIGURE 5. - Effect of NaOCI addition on gold 
 extraction. 
 
 with addition of 5 to 10 Ib/st NaOCl and 
 then increased gradually with larger 
 NaOCl additions. Gold extraction by sub- 
 sequent cyanidation reached 90 pet with 
 a 20-lb/st-NaOCl addition. The results 
 closely paralleled similar laboratory ex- 
 periments conducted on a different sample 
 of carbonaceous gold ore. 
 
 In other experiments, a reaction time 
 of 3 h (instead of 4 h) was shown to be 
 sufficient when 20 lb of lime per short 
 ton was used. It was also determined 
 that the optimum temperature for oxida- 
 tion with NaOCl lies between 50° and 
 60° C. 
 
 Oxidation by Chlorine Addition 
 
 Addition of chlorine to a pulp of fine- 
 ly ground carbonaceous ore and water was 
 investigated as a means of producing hy- 
 pochlorite oxidant in situ on an economi- 
 cal and easily controlled basis. As 
 chlorine is bubbled into the pulp, it re- 
 acts with water to form HOCl and 
 HCl. These products are buffered by the 
 calcareous gangue in the ore to form 
 Ca(0Cl)2 and CaCl2 . Oxidation of the 
 carbonaceous matter is thought to be ac- 
 complished by the hypochlorite ion. 
 
 It had been shown in the laboratory ex- 
 periments that addition of chlorine gas 
 at moderate rates to the agitated pulp 
 through a sintered-disk sparger resulted 
 in favorable oxidation without excessive 
 
 loss of chlorine gas. The chlorine re- 
 acted rapidly, and gold extraction from 
 the oxidized pulp by subsequent cyanida- 
 tion was shown to be largely independent 
 of the rate of chlorine addition. The 
 factor limiting the rate of chlorine ad- 
 dition appeared to be the amount of hypo- 
 chlorite remaining after oxidation. The 
 hypochlorite product reacted with the 
 carbonaceous matter as chlorine was added 
 during the initial stages of the oxida- 
 tion, and no hypochlorite could be de- 
 tected in the solution. The amount of 
 hypochlorite in solution increased with 
 time until the oxidation was completed. 
 It was determined that the oxidation can 
 be completed in as little as 4 h; how- 
 ever, at this relatively rapid rate of 
 chlorination, the Ca(0Cl)2 in solution 
 can increase to 1 pet or more in the 
 later stages of oxidation. If hypochlo- 
 rite is allowed to rise to a level that 
 cannot be consumed by the ore, it will 
 constjme cyanide in the cyanide circuit. 
 Therefore, it is desirable to limit the 
 Ca(0Cl)2 in solution to less than 0.2 pet 
 during chlorination. This resulted in a 
 chlorination time of up to 15 h at 24° to 
 30° C for some of the more refractory 
 samples, with an additional 5 h for the 
 ore to consume the residual hypochlorite. 
 The final concentration of Ca(0Cl)2 going 
 to cyanidation was less than 0.02 pet. 
 
 Pilot plant tests were run with three 
 55-gal drums connected in series, each 
 containing 275 lb of solids at 40 pet 
 solids. Chlorine was fed to each drum 
 through a 1/2-in-diam pipe extending to 
 the bottom of the drum. Oxidation was 
 conducted on a batch basis with the pulp 
 overflow from the third tank being con- 
 tinuously recirculated through the system 
 to give the desired chlorination time. 
 The near-optimum temperature of 24° to 
 30° C was maintained by heating the pulp. 
 The pH of the system remained at 6,6 dur- 
 ing chlorination without adjustment. 
 Figure 6 shows that gold extraction in- 
 creased rapidly with increasing chlorine 
 addition. The optimum extraction 92 pet, 
 was obtained using 40 Ib/st chlorine. 
 
 Scale-up experiments were conducted in 
 a 6- by 7-ft tank containing 7.5 st of 
 pulp at 40 pet solids. Chlorine gas was 
 bubbled in through four pipes submerged 
 
32 
 
 t3 100 
 
 80- 
 
 
 1 1 
 
 ^Q^ — ' ° 
 
 
 Y 
 
 
 ^^"-"-''^ Conditions: 
 
 
 
 
 Reaction time, 
 
 15 h 
 
 
 
 pH, 6.6 
 
 
 
 
 Temp, 24° to 3C 
 
 1 1 
 
 "C 
 
 
 Cathode 
 busbar 
 
 20 
 
 30 40 
 
 CHLORINE, Ib/st 
 
 50 
 
 FIGURE 6. - Effect of chlorine addition on gold 
 extraction. 
 
 5 ft into the tank. Chlorine was added 
 over an 8-h period at the rate of 3.5 lb/ 
 (sfh) initially, then decreased to 1.5 
 lb/(sfh) over the last 2.5 h. Gold 
 extractions of 89 pet (0.035 oz Au per 
 short ton of tails) were obtained using 
 38 Ib/st chlorine. These results paral- 
 leled those obtained in the pilot-plant- 
 scale experiments. 
 
 Electrooxidation of Carbonaceous Ores 
 
 In additional pilot plant studies, 
 electrolysis used on ore pulp prepared 
 from finely ground carbonaceous ore and 
 salt brine was investigated as a means of 
 generating oxidizing conditions in situ. 
 The cell used was a simple plate-type 
 arrangement consisting of graphite elec- 
 trodes 2-1/2 in wide, 3/4 in thick, and 
 30 in long (fig. 7). Identical graphite 
 cathodes and anodes were placed alter- 
 nately in a nonconductive holder with 
 1/2-in spacing, which allowed favorable 
 pulp flow through the system. 
 
 The pilot plant investigations deter- 
 mined the effect of temperature on gold 
 extractions, the effect of salt concen- 
 tration on gold extractions and power re- 
 quirements, and the effect of electrode 
 spacing at various salt concentrations 
 on power requirements. The data obtained 
 were similar to the data obtained in pre- 
 vious laboratory tests. Gold extractions 
 of 90 pet were obtained using 10-pct-salt 
 concentrations at 40° C, a current den- 
 sity of 0.67 A/in^ , and an electrode 
 spacing of 1/2 in. 
 
 Scale-up experiments were conducted on 
 a batch basis using a 6- by 7-ft agitator 
 tank capable of holding 7.5 st of pulp at 
 40 pet solids. Electrooxidation was ac- 
 complished with two banks of graphite 
 
 Pulp level 
 
 Anode 
 busbar 
 
 Electrodes 
 
 Ore slurry 
 
 30-50 pet 
 
 solids 
 
 FIGURE 7. - Plate-type electrode assembly for di- 
 rect electrolytic production of NaOCl in ore slurry. 
 
 90 
 
 o 
 
 Q. 
 
 g 80 
 
 < 
 
 I- 
 ^ 70 
 
 Q 
 _1 
 O 
 CD 
 
 60 
 
 1 
 
 / Conditions: 
 
 2,800 A at 5.2 V 
 
 =t.*- — 
 
 1 
 
 Temp, 40° C 
 
 Salt cone, 10 pet 
 
 1 1 1 
 
 
 20 30 40 50 60 
 
 ENERGY CONSUMPTION, kW-h/st 
 
 70 
 
 FIGURE 8. - Gold extraction and energy consump- 
 tion for large-scale experiments. 
 
 electrodes used in parallel, each con- 
 taining 12 anodes and 11 cathodes. Tests 
 were conducted at 40° C, with a 2,800-A 
 current and a 10-pct-salt concentration 
 in the pulp solution. Oxidized pulps 
 were not cyanided in the pilot plant, but 
 samples were taken at 1-h intervals and 
 cyanided on a bench scale. Figure 8 
 shows that gold extraction increased with 
 electrolysis time, reaching a maximum ex- 
 traction of 89.4 pet, which corresponded 
 to 22 h of electrolysis. Data obtained 
 
33 
 
 from the larger experiments indicated 
 that scale-up to commercial plants should 
 not present any serious problems. 
 
 The pilot plant studies were a coopera- 
 tive effort of the Newmont Mining Corp.'s 
 Carlin Gold Mine and the Bureau of Mines. 
 Experiments were conducted at the Carlin 
 
 Gold Mine and at the Bureau's Reno (NV) 
 Research Center. Based on the results of 
 the testing program, the Carlin Mine in- 
 stalled a full-scale facility using chlo- 
 rine to treat carbonaceous ores prior to 
 cyanidation. 
 
 CONCLUSIONS 
 
 Treatment of carbonaceous gold ores 
 from north-central Nevada with NaOCl or 
 chlorine, or by electrolytic oxidation 
 resulted in favorable gold recovery by 
 subsequent cyanidation. The gold that is 
 complexed by humic acid- type compounds 
 was liberated, and the adsorptive proper- 
 ties of the ore were passivated by the 
 oxidation treatment. Gold extractions in 
 the 90-pct range were obtained on several 
 
 tonnage ore samples from the Carlin Gold 
 Mine. Equivalent metallurgical perfoirm- 
 ance, compatible with present plant prac- 
 tice, was obtained with NaOCl, chlorina- 
 tion, and electrooxidation. The choice 
 of one of these methods is therefore dic- 
 tated by economic conditions such as the 
 cost and availability of electric power, 
 NaOCl, and chlorine. 
 
 REFERENCES 
 
 1, Leaver, E. S., and J. A. Woolf. 
 Re-Treatment of Mother Lode (California) 
 Carbonaceous Slime Tailings. BuMines 
 Tech. Paper 481, 1930, 20 pp. 
 
 2, Hedley, N, , and H, Tabachnick, 
 Chemistry of Cyanidation, Mineral Dress- 
 ing Notes, No. 23. Am. Cyanamid Co., New 
 York, June 1958, 54 pp. 
 
 3, Radtke, A, S,, and B, J, Scheiner, 
 Studies of Hydrothermal Gold Deposition 
 (I), Carlin Gold Deposit, Nevada: The 
 Role of Carbonaceous Material in Gold 
 Deposition, Econ, Geol, , v, 65, 1970, 
 pp, 87-102, 
 
 4, Radtke, A, S,, C. Heropoulos, B, P, 
 Fabbi, B. J, Scheiner, and M, Essington. 
 Data on Major and Minor Elements in Host 
 Rocks and Ores, Carlin Gold Deposit, Ne- 
 vada. Econ. Geol., v. 67, 1972, pp. 975- 
 978. 
 
 5, Hausen, D. M. , and P. F. Kerr. 
 Fine Gold Occurrence at Carlin, Nevada, 
 Sec, in Ore Deposits of the United 
 States, 1933-1967, ed, by J, D, Ridge, 
 AIME, V, 1, 1968, pp, 908-940, 
 
 6, Scheiner, B, J. , R. E. Lindstrom, 
 and T, A. Henrie. Investigation of Oxi- 
 dation Systems for Improving Gold Recov- 
 ery From Carbonaceous Materials. BuMines 
 TPR 2, July 1968, 8 pp. 
 
 and C. M. Frost, 
 Byproduct, Bu- 
 
 pp, 
 R, E. Lindstrom, 
 
 7, Scheiner, B, J,, R, E, Lindstrom, 
 and T, A, Henrie, Oxidation Process 
 for Improving Gold Recovery From Carbon- 
 Bearing Gold Ores, BuMines RI 7573, 
 1971, 14 pp, 
 
 8, Fowkes, W, W, , 
 Leonardite: A Lignite 
 Mines RI 5611, 1960, 12 
 
 9, Scheiner, B, J,, 
 and T, A, Henrie, Process for Recovery 
 of Gold From Carbonaceous Ores, U.S, 
 Pat. 3,574,600, Apr. 13, 1971. 
 
 10. Scheiner, B. J., R. E. Lindstrom, 
 and T. A. Henrie. Electrolytic Oxidation 
 of Carbonaceous Ores for Improving Gold 
 Recovery. BuMines TPR 8, 1969, 12 pp. 
 
 11. . Recovery of Gold From Car- 
 bonaceous Ores. U.S. Pat. 3,639,925, 
 Feb. 8, 1972. 
 
 12. Scheiner, B, J,, R. E, Lind- 
 strom, W. J. Guay, and D, G, Peterson, 
 Extraction of Gold From Carbonaceous 
 Ores: Pilot Plant Studies, BuMines RI 
 7597, 1972, 20 pp, 
 
 13. Scheiner, B, J,, R, E. Lindstrom, 
 and T, A, Henrie. Processing Refractory 
 Carbonaceous Ores for Gold Recovery. J. 
 Met., V. 23, No. 3, 1971, pp. 37-40. 
 
34 
 
 CARBON ADSORPTION-DESORPTION 
 By J. A. Eisele"! 
 
 ABSTRACT 
 
 The Bureau of Mines has developed 
 three methods for stripping gold and 
 silver from activated carbon that enable 
 the carbon to be reused many times. 
 These three desorption techniques are to 
 strip the precious metals with either 
 (1) boiling NaOH-NaCN solution at atmos- 
 pheric pressure; (2) NaOH-NaCN solution 
 
 at temperatures above boling, in pressure 
 vessels; or (3) an NaOH-NaCN solution 
 containing ethyl alcohol at temperatures 
 below boiling. All three methods are 
 widely applied in industry. This paper 
 describes the three methods and their 
 development. 
 
 INTRODUCTION 
 
 The introduction of cyanide leaching in 
 the 1890' s revolutionized the processing 
 of gold and silver ores. Since that 
 time, the standard hydrometallurgical 
 process to recover gold and silver from 
 disseminated ores has become cyanide 
 leaching in a countercurrent decantation 
 (CCD) plant (U.2 The operation of a 
 conventional CCD plant (fig. 1) consists 
 of the following steps: 
 
 1. The ore is crushed and ground. 
 
 2. During grinding, the ore is slur- 
 ried with dilute basic cyanide solution 
 typically containing 0.1 pet NaCN and 
 
 'Supervisory chemical engineer, Reno 
 Research Center, Bureau of Mines, Reno, 
 NV. 
 
 ^Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
 c'ao" ^^°o'^' K-l Comminution I —\ Leaching lonk I I JLeoctiing lonk 2 
 
 ]^ 
 
 t^ 
 
 ^7>^ 
 
 Precipilalion \-^ Zinc du5 
 
 Gold-silver sponge 
 
 FIGURE 1. - Flow diagram of countercurrent decan- 
 tation cyanide leaching plant. 
 
 enough CaO to maintain a pH range of 10 
 to 11. 
 
 3. The slurry is agitated in a series 
 of leaching tanks to give a total leach- 
 ing time of 12 to 48 h, depending on the 
 characteristics of the ore. 
 
 4. The slurry is thickened and 
 washed countercurrently in a series of 
 thickeners. 
 
 5. The pregnant solution from the 
 first thickener is clarified by filtra- 
 tion and deaerated by vacuum. 
 
 6. Zinc powder is added to precipitate 
 the gold and/or silver. 
 
 7. The gold-silver-zinc sponge is fil- 
 tered and refined to bullion, 
 
 8. The barren cyanide solution is re- 
 turned to the washing-leaching circuit, 
 and makeup cyanide and base are added. 
 
 Gold and silver recovery from pregnant 
 solutions by zinc precipitation can pre- 
 sent problems. Ideally, the pregnant so- 
 lution should be clear before precipita- 
 tion, and it, must be deaerated. This can 
 be difficult to achieve with slimy ores. 
 Also, precipitation is an inefficient 
 method for recovering gold from dilute 
 solutions. Different forms of carbon, 
 especially activated carbon, have long 
 been known to be good adsorbers of gold 
 and silver cyanides from solution. How- 
 ever, since the only known means to re- 
 cover the gold and silver from the carbon 
 was to burn it, carbon was not used ex- 
 tensively. In a few cases, where slimy 
 ores were being treated, gold adsorption 
 on carbon was the preferred method for 
 
35 
 
 recovering the precious metal values be- 
 cause the clarification step was elimi- 
 nated. The precious-metal-loaded carbon 
 
 was recovered from the 
 ing or flotation. 
 
 slurry by screen- 
 
 USE OF BOILING NaOH-NaCN SOLUTION 
 
 In the late 1940' s, Bureau of Mines re- 
 searchers were looking for ways to strip 
 the precious metals from loaded carbon to 
 enable the carbon to be reused a number 
 of times, A World War II surplus of ac- 
 tivated carbon manufactured from fruit 
 pits was available at prices that made 
 using carbon cheaper than using zinc, A 
 means for desorbing and recycling the 
 carbon would make carbon adsorption a 
 very economical process for the recov- 
 ery of gold. In 1950, an alkaline Na2S 
 stripping method was described that 
 eluted the gold, but not the silver, from 
 carbon (2^), As most ores contain some 
 silver, and silver in some ores is the 
 primary value, the sulfide stripping 
 method was not satisfactory because the 
 carbon would eventually be loaded with 
 unstrippable silver. In 1952, Bureau re- 
 searchers reported a method for desorbing 
 gold and silver from loaded carbon (3), 
 Precious metals were stripped by contact- 
 ing the carbon with a boiling NaOH-NaCN 
 solution. With each pass of the solution 
 through the carbon bed, some of the gold 
 and silver was removed. After being 
 passed through the carbon bed, the solu- 
 tion entered an electrolysis cell, where 
 the remaining gold and silver were depos- 
 ited on the cathode. The barren solution 
 was then reheated and recycled through 
 the carbon. A 1 pet NaOH-0,1 pet NaCN 
 solution at its boiling point desorbed 
 more than 90 pet of the gold and silver 
 in 4 to 6 h. The carbon could be reused 
 up to 10 times without losing significant 
 activity. 
 
 Carbon adsorption-desorption-electro- 
 winning permitted gold and silver 
 
 recovery from slimy ores or the slimy 
 portion of ores by eliminating the re- 
 quirement of a clarified solution. This 
 became the foundation for two important 
 developments in gold and silver ore pro- 
 cessing: carbon-in-pulp (CIP) cyanida- 
 tion and heap leaching. The cylindrical 
 electrolytic cell originally described 
 (2^) is still used for electrowinning and 
 is commonly referred to as a "Zadra" 
 cell, even though many units in commer- 
 cial use have been modified. Rectangular 
 electrowinning cells are also used, to 
 make better use of floor space than cyl- 
 indrical cells. The NaOH-NaCN electro- 
 lyte and steel wool cathode remain common 
 to all gold-silver electrowinning cells. 
 The Bureau's desorption-electrowinning 
 process, known as the Zadra process, has 
 been utilized in cyanide milling for more 
 than 30 yr. 
 
 Although gold and silver were eluted 
 from the fruit pit carbon in 4 to 6 h 
 in pilot-scale tests, this was not the 
 general case. Commercial practice, which 
 had adopted harder carbons made from 
 coconut shells, showed that 24 to 48 h 
 was required to desorb more than 90 pet 
 of the precious metals using alka- 
 line cyanide solution heated to boil- 
 ing. Although this was better than not 
 being able to strip the carbon, the long 
 stripping time was undesirable. Bureau 
 research was directed toward ways to 
 decrease the stripping time. Two meth- 
 ods, pressure stripping and alkaline- 
 alcohol stripping, were developed; both 
 greatly decreased the desorption time. 
 
 PRESSURE STRIPPING AND ALKALINE-ALCOHOL STRIPPING 
 
 In 1973, Bureau researchers showed that 
 by using a pressure vessel and increasing 
 the temperature of the stripping solu- 
 tion, gold could be stripped from carbon 
 in 2 to 6 h (4^) , The loaded carbon was 
 conditioned with caustic-cyanide solution 
 at 90° C and eluted with water at 150° C, 
 
 One advantage was that consumption of 
 cyanide and caustic was less with heated 
 pressure stripping than it was using pro- 
 longed ambient pressure stripping. The 
 stripping solution was cooled to 90° C 
 and the gold recovered by electrowinning. 
 
36 
 
 In the alkaline-alcohol stripping meth- 
 od, developed by Bureau researchers in 
 1976, ambient pressure was used, but the 
 modified stripping solution contained 20 
 pet ethanol in addition to the alkaline 
 cyanide (_5 ) . At 80° C, gold and silver 
 were desorbed in 6 h. A concurrent de- 
 velopment was the separation of gold from 
 silver by precipitating silver as a sul- 
 fide (6^). The separation of silver as a 
 sulfide takes place after desorption from 
 
 the carbon and results in a purer gold 
 bullion. For pregnant solutions contain- 
 ing considerable amounts of silver, the 
 preferable sequence is to precipitate the 
 silver before loading the carbon (_7 ) • 
 A large carbon inventory is avoided by 
 maintaining capacity for only gold ad- 
 sorption. The Ag2S precipitate can be 
 smelted to a silver bullion. The gold in 
 the filtrate from silver precipitation is 
 adsorbed, desorbed, and electrowon. 
 
 SUMMARY 
 
 Pressure stripping, alkaline-alcohol 
 stripping, and ambient pressure stripping 
 are all used by industry to strip gold 
 and silver from activated carbon. Using 
 
 any of these three methods, the carbon 
 can be reused many times. The preference 
 of the mill operator is the determining 
 factor in choosing the stripping system. 
 
 REFERENCES 
 
 1. McQuiston, F. W. , Jr., and R. S. 
 Shoemaker. Gold and Silver Cyanidation 
 Plant Practice. AIME, v. 1, 187 pp., 
 1975; V. 2, 263 pp., 1980. 
 
 2. Zadra, J. B. A Process for the Re- 
 covery of Gold From Activated Carbon by 
 Leaching and Electrolysis. BuMines RI 
 4672, 1950, 47 pp. 
 
 3. Zadra, J. B., A. L. Engel, and 
 H. J. Heinen. Process for Recovering 
 Gold and Silver From Activated Carbon by 
 Leaching and Electrolysis. BuMines RI 
 4843, 1952, 32 pp. 
 
 4. Ross, J. R. , H. B. Salisbury, and 
 G. M. Potter. Pressure Stripping Gold 
 From Activated Carbon. Pres. at Soc. 
 Min. Eng. AIME Annu. Conf., Chicago, IL, 
 Feb. 26-Mar. 1, 1973, 15 pp.; available 
 
 upon request from Jean Beckstead, Bureau 
 of Mines, Salt Lake City, UT. 
 
 5. Heinen, H. J. , D. G. Peterson, and 
 R. E. Lindstrom. Gold Desprption From 
 Activated Carbon With Alkaline Alcohol 
 Solutions. Ch. 33 in World Mining and 
 Metals Technology, ed. by A. Weiss (Proc. 
 Joint Min. and Metall. Inst, of Japan- 
 AIME Meeting, Denver, CO, Sept. 1-3, 
 1976). AIME, 1976, pp. 551-564. 
 
 6. . Processing Gold Ores Using 
 
 Heap Leach-Carbon Adsorption Methods. 
 BuMines IC 8770, 1978, 21 pp. 
 
 7. . Silver Extraction From Mar- 
 ginal Resources. Pres. at 104th TMS-AIME 
 Annu. Meeting, New York, Feb. 16-20, 
 1975, 14 pp.; available upon request from 
 J. A. Eisele, Bureau of Mines, Reno, NV. 
 
37 
 
 HEAP LEACHING 
 By J. A. Eiselel 
 
 ABSTRACT 
 
 The Bureau of Mines began investigat- 
 ing heap leaching in 1969 in an effort 
 to provide a low-cost means for recover- 
 ing precious metals from ores that were 
 too low in grade to be economically pro- 
 cessed by conventional cyanide technol- 
 ogy* By 1979, the Bureau had developed 
 
 an agglomeration pretreatment method that 
 permitted heap leaching to be applied to 
 clayey and finely divided materials. To- 
 day, heap leaching is widely used by the 
 mining industry. This paper briefly de- 
 scribes and discusses heap leaching and 
 agglomeration pretreatment. 
 
 INTRODUCTION 
 
 Heap leaching was first used on oxide 
 copper ores and uranium ores (JL^).^ It 
 had the advantages of very low capital 
 cost, low operating costs, and opera- 
 tional flexibility. Bureau researchers 
 proposed heap leaching in 1969 as a low- 
 cost means for recovering gold values 
 from disseminated gold ores with porous 
 gangue, mine stripping waste, and submar- 
 ginal ores (2-4). Heap leaching has 
 
 since become an important factor in pre- 
 cious metals recovery because it permits 
 utilization of lean ores and wastes that 
 cannot be economically processed by con- 
 ventional agitation cyanidation. Fifteen 
 years after the method's introduction, 
 there are between 50 and 100 commercial 
 gold and silver heap leaching operations, 
 ranging in size from 10 st/wk to 10,000 
 st/d. 
 
 LEACHING 
 
 In heap leaching, the crushed ore mate- 
 rial is piled in heaps on impervious 
 pads. A dilute alkaline-cyanide solution 
 is distributed on top of the heap by a 
 sprinkling system. The solution perco- 
 lates through the heap and drains from 
 the impervious pad. The pregnant gold 
 solution from the heap typically contains 
 from 1 to 3 ppm Au.- 
 
 For many operations, precipitation of 
 the metal values with zinc is not the 
 most efficient recovery method. Another 
 method, carbon adsorption, is very effi- 
 cient for recovering of metals from di- 
 lute solutions. The pregnant solution is 
 passed through a series of columns con- 
 taining activated carbon, and the gold is 
 adsorbed. The resulting barren solution 
 
 -I , 
 
 'Supervisory chemical engineer, Reno 
 
 Research Center, Bureau of Mines, Reno, 
 NV. 
 
 ■^Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
 is fortified with reagents and recycled 
 to the heap. Leaching continues until 
 the gold extraction is completed. The 
 gold-loaded carbon is stripped by carbon 
 adsorption or zinc precipitation. Some 
 pregnant solutions from silver heap 
 leaching contain enough silver, typically 
 10 to 20 ppm Ag, that zinc precipitation 
 can be used to recover the silver. 
 
 For heap leaching to be feasible, .the 
 ore must be porous and permeable to the 
 leaching solution. Since the ore is not 
 finely ground, the cyanide ions must dif- 
 fuse through the host rock to dissolve a 
 gold particle. The dissolved gold must 
 diffuse outward. This requires leaching 
 cycles weeks or months long, and may re- 
 sult in dissolved gold cyanide complexes 
 that are not completely washed from the 
 heap. Runoff from abandoned heaps is not 
 discharged to surface or groundwater 
 sources until the effluent from the spent 
 heap is free of cyanide. A flow diagram 
 of a typical heap leaching operation is 
 shown in figure 1. 
 
38 
 
 PRETREATMENT 
 
 Although application of the heap leach- 
 ing technique to gold and silver ores 
 permitted development of many low-grade 
 and/or small properties that otherwise 
 would not have been exploitable, some 
 ores were untreatable by heap leaching. 
 This was due to two conditions: (1) The 
 ore contained clay, which swelled on con- 
 tact with the leaching solution, blocked 
 the voids in the heap, and thereby pre- 
 vented solution flow; and (2) the ore, 
 after crushing to the liberation size for 
 gold-silver extraction, generated an un- 
 usually large amount of fines (minus 200 
 mesh) that were washed into the voids by 
 
 Makeup NaCN, 
 CaO, HgO 
 
 _.i__ 
 
 Ore- 
 
 
 Heap leaching 
 
 --i- 
 
 Car 
 
 bon 
 
 
 
 
 
 
 adsorption 
 
 . 
 
 
 
 
 
 
 ' 
 
 
 
 
 Carbon 
 desorption 
 
 Ttiermal 
 reactivation 
 
 r -*■ 
 
 
 ' 
 
 keup No 
 
 1 
 
 i 
 
 t 
 
 
 1 
 
 
 1 
 
 Electrowinning 
 
 
 
 
 
 
 
 Refining 
 
 
 Gold-silver bullion 
 FIGURE 1. - Flow diagram of heap leaching. 
 
 the percolating leaching solution, caus- 
 ing channeling and incomplete leaching 
 of the gold and silver from the heap 
 material. 
 
 In 1979, Bureau researchers published a 
 report describing an agglomeration method 
 that was successful in overcoming these 
 problems (_5-^) , Agglomeration as a pre- 
 treatment for heap leaching consists of 
 (1) mixing the crushed ore with portland 
 cement, which acts as a binding agent, 
 and lime to provide alkalinity; (2^) wet- 
 ting the mixture evenly with solution, 
 which may contain cyanide to start leach- 
 ing before the heap is built; and (3) me- 
 chanical tumbling of the mixture so 
 the fine particles adhere to the larger 
 particles. Several hours of aging are 
 needed for the cement to bond the parti- 
 cles. When stable bonds are formed, the 
 agglomerates are very durable and resist- 
 ant to degradation. This simple pre- 
 treatment has increased the flow of 
 some ores through the columns as much as 
 6,000-fold and, in actual heaps, has 
 decreased the leaching cycle to days in- 
 stead of weeks. It is estimated that 
 half of the heap leaching operations use 
 some type of agglomeration pretreatment 
 (7^), Agglomeration pretreatment can also 
 be applied to finely divided material 
 such as tailings or finely ground ore 
 (8^) , The conditions required to form 
 good agglomerates are more rigorous, but 
 when the criteria are met, it is possible 
 to process finely divided low-grade and/ 
 or small resources. 
 
REFERENCES 
 
 39 
 
 1. Potter, G. M. Design Factors for 
 Heap Leaching Operations, Min, Eng. , 
 Mar. 1981, pp 277-281. 
 
 2. Heinen, H. J., and B. Porter. Ex- 
 perimental Leaching of Gold From Mine 
 Waste. BuMines RI 7250, 1969, 5 pp. 
 
 3. Merwin, R. W. , G. M. Potter, and 
 H. J. Heinen. Heap Leaching of Gold Ores 
 in Northwestern Nevada. (Pres. at AIME 
 Annu. Meeting, Washington, DC, Feb. 16- 
 20, 1969). Soc. Min. Eng. AIME preprint 
 69-AS-79, 1969, 15 pp. 
 
 4. Potter, G. M. Recovering Gold From 
 Stripping Waste and Ore by Percolation 
 Cyanide Leaching. BuMines TPR 20, 1969, 
 5 pp. 
 
 5. Heinen, H. J., G. E. McClel- 
 land, and R. E. Lindstrom. Enhancing 
 
 Percolation Rates in Heap Leaching of 
 Gold-Silver Ores. BuMines RI 8388, 1979, 
 20 pp. 
 
 6. McClelland, G. E., and J. A. 
 Eisele. Improvements in Heap Leaching To 
 Recover Silver and Gold From Low-Grade 
 Resources. BuMines RI 8612, 1982, 26 pp. 
 
 7. McClelland, G. E., D. L. Pool, and 
 J. A. Eisele. Agglomeration-Heap Leach- 
 ing Operations in the Precious Metals In- 
 dustry. BuMines IC 8945, 1983, 16 pp. 
 
 8. McClelland, G. E., D. L. Pool, 
 A. H. Hunt, and J. A. Eisele. Agglomera- 
 tion and Heap Leaching of Finely Ground 
 Precious-Metal-Bearing Tailings. BuMines 
 IC 9034, 1985, 11 pp. 
 
40 
 
 THE CARBON-IN-PULP PROCESS 
 By Stephen D. Hill'' 
 
 ABSTRACT 
 
 This paper briefly reviews the develop- 
 ment of the carbon-in-pulp (CIP) process 
 for recovering gold and silver from cya- 
 nide leach solutions. A brief history 
 and a description of the various steps 
 in the process are presented. Reference 
 
 is made to published research findings, 
 from both the Bureau of Mines and other 
 sources, that contributed significantly 
 to commercial development and utilization 
 of the process. 
 
 INTRODUCTION 
 
 The application of activated carbon to 
 recover gold and silver from cyanide 
 leach solutions was patented as early as 
 1894 (J_).2 T. G. Chapman, of the Univer- 
 sity of Arizona, is credited with the 
 original development of CIP processing in 
 the late 1930' s. In Chapman's system, 
 the dissolved gold was adsorbed by finely 
 ground activated charcoal. Subsequently, 
 the charcoal was separated from the pulp 
 by flotation, and gold was recovered from 
 the charcoal by burning or smelting (2). 
 
 In an alternative system. Chapman used 
 larger particles of activated carbon, 
 particles that were much coarser than the 
 ground ore, as the gold adsorbent. The 
 carbon, enclosed in a cylindrical screen 
 basket, was circulated and rotated within 
 the leach pulp. A concentrated gold so- 
 lution was obtained by stripping the 
 gold-loaded carbon with hot cyanide solu- 
 tion (_3 ) . Bureau of Mines researchers 
 further modified the Chapman system in 
 the early 1950' s by adding an electro- 
 lytic cell for continuous removal of gold 
 from the hot strip liquor (4). 
 
 A small commercial CIP plant was op- 
 erated from 1954 to 1960 at the Golden 
 Cycle Corp.'s Carlton Mill in Cripple 
 Creek, CO. In that operation, coarse 
 carbon in screen baskets was loaded 
 
 'Research Director, Salt Lake City Re- 
 search Center, Salt Lake City, UT. 
 
 ^Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
 with gold 
 stripped 
 method (5) 
 In 1970, 
 continued 
 
 to about 40 tr oz/st Au, then 
 and recycled by the Zadra 
 
 the Homestake Mining Co. dis- 
 mercury amalgamation in its 
 milling plant in Lead, SD, because of 
 downstream pollution. Since the ore had 
 a high slime content, Homestake' s diffi- 
 culties provided an opportunity for the 
 Bureau to test its own concepts for im- 
 proved gold extraction, loading, and car- 
 bon handling in a modified cyanide CIP 
 system. Research in these areas was per- 
 formed under a cooperative agreement be- 
 tween the Bureau and Homestake. At the 
 time, conventional plants used a costly 
 system for processing low-grade, lode 
 gold ores that contained finely dissemi- 
 nated gold. Considerable capital invest- 
 ment was required for fine grinding, 
 agitation leaching in cyanide solution, 
 countercurrent decantation in thickeners 
 for solid-liquid separation, residue 
 washing, clarification of pregnant solu- 
 tion by filtering, vacuum deaeration, and 
 precipitation of the gold by reacting the 
 solution with zinc powder. 
 
 The alternative system, proposed by 
 the Bureau, optimized the following oper- 
 ations: pulp leaching, CIP loading, 
 carbon separation, stripping, electrowin- 
 ning, and carbon regeneration. The inno- 
 vative system was perfected and demon- 
 strated at Homestake in a miniplant that 
 handled 2 st/d of slime pulp. By extrap- 
 olation of the design from the mini- 
 plant, Homestake was able to start its 
 
41 
 
 commercial CIP plant, treating 2,000 st/d 
 slime, in 1973 (6^). Bureau metallurgists 
 reported on the Homestake project and the 
 new CIP process at the American Mining 
 Congress Mining Convention in Denver, CO, 
 the same year (7^) . 
 
 The Bureau subsequently developed the 
 carbon pressure stripping process, and 
 refinements were made in a new design for 
 the electrowinning cell. Cost data pre- 
 sented in 1974 clearly showed that the 
 CIP system was superior to the conven- 
 tional countercurrent decantation system 
 (8^). The MINTEK organization in Johan- 
 nesburg, Republic of South Africa, began 
 work on CIP in 1976, and many significant 
 developments have resulted from its in- 
 tensive efforts (9) . 
 
 Other CIP plants patterned after the 
 Bureau process include Duval Corp.'s Bat- 
 tle Mountain, NV, operation in 1978 (10); 
 Pinson Mining Co.'s Winnemucca, NV, plant 
 in 1980 (11); and Freeport Gold Co.'s 
 operation near Elko, NV, in 1981, the 
 largest plant of its kind in the United 
 States (12) . Getty Mining Co., Mercur, 
 UT, started a carbon-in-leach (CIL) 
 plant, an offshoot from the CIP process, 
 in 1983 (13). Today, the CIP process 
 is the standard of the gold industry 
 throughout the world, with 30 plants in 
 North America, 10 plants in Australia, 
 and approximately 20 plants in the Repub- 
 lic of South Africa using the process 
 (24-J^) . 
 
 LEACHING 
 
 The CIP process, shown in figure 1, 
 Includes cyanide agitation leaching, 
 countercurrent carbon-pulp contact, and 
 carbon-pulp separation. The leaching 
 step of the process, usually accomplished 
 in a multistage system, comprises sev- 
 eral tanks; four to six stages are typi- 
 cal. The ore, normally ground to minus 
 100 mesh or finer, is leached in a thick 
 slurry of about 45 to 50 pet solids. 
 Lime or caustic is added to maintain 
 protective alkalinity. To speed up gold 
 dissolution, the pulp is vigorously aer- 
 ated and agitated during leaching. The 
 retention time necessary for complete 
 
 Gold ore 
 
 Crushing 
 
 Grinding -• — NaCN plus lime 
 
 45-50 pet solids pulp, minus 100 mesh or finer 
 
 Leach agitators 
 
 Stripping »• Electrowinning ►Gold 
 
 Carbon reactivation 
 
 Makeup 
 
 minus 10- plus 20-mesh 
 
 activated c arbo n 
 
 1 
 
 FIGURE 1.- CIP process. 
 
 dissolution of the gold depends greatly 
 on the particle size of the gold, but 2 
 to 24 h is typical. Other factors that 
 influence the rate of dissolution include 
 particle size of the ore, clay content, 
 pulp density, pH, cyanide strength, tem- 
 perature, and slurry viscosity. 
 
 The Dorr agitator, with a slow-speed 
 center sweep and either peripheral or 
 center-column airlifts, has worked well 
 on finely ground ores. Pachuca tanks 
 also have been successful; they may be 
 capable of handling a coarser feed than 
 the traditional Dorr tanks, A draft 
 tube-type agitation tank has also been 
 used successfully. The type of vessel 
 most suitable for CIP processing depends 
 on the type of ore being treated and the 
 prevailing operating conditions. In most 
 cases, a deep tank with turbine-type me- 
 chanical agitators and low tip speeds is 
 preferred (17), Dissolution of gold and 
 silver is essentially completed before 
 the leached pulp moves to the CIP adsorp- 
 tion circuit. 
 
42 
 
 ADSORPTION 
 
 CARBON SEPARATION 
 
 The CIP adsorption circuit is a cascade 
 of agitated vessels through which the 
 pulp flows by gravity. In each vessel, 
 the pulp is contacted with carbon gran- 
 ules that preferentially adsorb gold and 
 silver from the solution as the pulp 
 overflows from one vessel into the next. 
 Periodically, a portion of the carbon in- 
 ventory is transferred up the cascade to 
 the next vessel by airlifting a quantity 
 of carbon-bearing pulp. The transfer 
 system is abrasive; therefore, coconut 
 shell activated carbon is generally used 
 because it is resistant to abrasion. 
 There is some variation in the size of 
 carbon particle used. The only strict 
 criterion for carbon size is that parti- 
 cles be large enough to allow easy sepa- 
 ration from the pulp; a minus 6- plus 16- 
 mesh carbon size is quite common. 
 
 The optimum loading of gold and silver 
 on the carbon is generally a matter of 
 economics and is very dependent on metal 
 prices. Consideration must be given to 
 the ore grade, solution value, gold in- 
 ventory, and security. Frequent strip- 
 ping and handling can result in lower 
 loading, so most large operations usually 
 load the carbon to 200 to 400 tr oz/st Au 
 before the carbon is removed from the 
 first tank in the cascade. 
 
 Several factors influence the adsorp- 
 tion capacity of gold onto activated car- 
 bon from cyanide solutions. Some of the 
 more important are the ionic strength and 
 pH value of the solution, temperature, 
 the presence of competing metal ions and 
 poisons, and the nature of the carbon. 
 In addition, a number of factors influ- 
 ence the extraction rate of gold cyanide 
 from activated carbon in the stirred 
 tanks, the most important of which are 
 the mixing efficiency in the tanks, pulp 
 viscosity, and carbon granule sizes (18) , 
 
 The Bureau of Mines recommends that ad- 
 sorption efficiencies be determined by 
 conducting laboratory CIP batch tests to 
 obtain adsorption rate data and establish 
 equilibriim isotherms (19-20) . These ad- 
 sorption rates and equilibrium curves can 
 then be used to design the optimum ad- 
 sorption system. 
 
 Methods of separating the loaded carbon 
 particles from the gold-depleted pulp are 
 numerous and widely varied. In early CIP 
 circuits, screen baskets or external vi- 
 brating screens were used extensively. A 
 vibrating screen with a minus 20- plus 
 24-mesh stainless steel square mesh deck 
 has proven to be reliable. Such screens 
 have worked well at a number of plants 
 without malfunctioning. 
 
 A problem of carbon fines being gen- 
 erated caused operators to look for im- 
 proved methods that would save power, re- 
 duce fine carbon generation, and reduce 
 capital costs. The new concepts that re- 
 sulted include either a peripheral screen 
 or a submerged launder-type screen that 
 fits across the top of the CIP tank. 
 Typically there are two or three such 
 launders on each tank. Each launder has 
 two or three removable screens set in the 
 side panels, which may be vibrated for 
 carbon transfer, if necessary, 
 
 CARBON-IN-LEACH 
 
 Leaching generally requires a much 
 longer pulp residence time than adsorp- 
 tion. Consequently, it is possible to 
 reduce the equipment requirement in the 
 CIP circuit by using the leach vessels 
 for both cyanidation and adsorption si- 
 multaneously, thus eliminating the need 
 for a separate adsorption cascade. Such 
 a system is called a carbon-in-leach 
 (CIL) circuit (21) . Because the acti- 
 vated carbon adsorbs gold and silver from 
 solution, not from the ore, there are ad- 
 vantages in partially preleaching the 
 ore before adsorption starts. Leaching 
 can go on in the presence of carbon that 
 is moving countercurrent to the flow of 
 pulp, similar to the operation of a stan- 
 dard CIP circuit, 
 
 RECOVERY OF GOLD AND SILVER 
 
 The activated carbon, loaded with 
 gold and/or silver, may be shipped to a 
 smelter to recover the precious metal 
 values. Such a procedure is sometimes 
 used, particularly for small mines whose 
 
capital is limited and reserves are 
 small. However, for better economy in 
 larger operations, the carbon is prefer- 
 ably stripped, reactivated, and re- 
 used in the process. Details of carbon 
 
 43 
 
 stripping, regeneration, and recovery of 
 precious metals by electrowinning from 
 strip solution are discussed in other pa- 
 pers in this Information Circular. 
 
 REFERENCES 
 
 1. Johnson, W. D. Abstraction of 
 Gold and Silver From Their Solutions in 
 Potassium Cyanide. U.S. Pat. 533,260, 
 May 18, 1894. 
 
 2. Chapman, T. G. Cyanidation of 
 Gold Bearing Ores. U.S. Pat. 2,147,009, 
 Sept. 22, 1939. 
 
 3. Crabtree, E. H. , Jr., V. W. Win- 
 ters, and T. G. Chapman. Developments 
 in the Application of Activated Carbon 
 to Cyanidation. Metall. Trans., v. 187, 
 Feb. 1950, pp. 217-222. 
 
 4. Zadra, J. B., A. L. Engel, and 
 H. J. Heinen. Process for Recovering 
 Gold and Silver From Activated Carbon 
 by Leaching and Electrolysis. BuMines 
 RI 4843, 1952, 32 pp. 
 
 5. Seeton, D. A. A Review of Carbon 
 Cyanidization. Min. Mag., v. 51, July 
 1961, pp. 13-15. 
 
 6. Hall, K. B. Homestake Uses Car- 
 bon-in-Pulp To Recover Gold From Slimes. 
 World Min., v. 27, No. 12, 1974, p. 44. 
 
 7. Potter, G. M. , and H. B. Salis- 
 bury. Innovations in Gold Metallurgy. 
 (Pres, at Am. Min. Congr. Min. Conv. and 
 Environ. Show, Denver, CO, Sept. 9-12, 
 1973.) BuMines preprint (Salt Lake City, 
 UT), 1973, 12 pp. 
 
 8. Rosenbaum, J. B. Minerals Extrac- 
 tion and Processing: New Developments. 
 Science, v. 191, 1976, pp. 720-723. 
 
 9. Laxen, P. A., G. S. M, Becker, and 
 R. Rubin. Developments in the Applica- 
 tion of Carbon-in-Pulp to the Recovery of 
 Gold From South African Ores. J. S. Afr. 
 Inst. Min. & Metall., v. 79, June 1979, 
 pp. 315-326. 
 
 10. Jackson, D. (ed.). How Duval 
 Transformed Its Battle Mountain Proper- 
 ties From Copper to Gold Production. 
 Eng. and Min. J., v. 183, No. 10, 1982, 
 pp. 95-99. 
 
 11. Mining Magazine. Pinson, Nevada — 
 New Open Pit Gold Mine. V. 145, July 
 1981, p. 5. 
 
 12. Jackson, D. (ed.). Jerritt Canyon 
 Project. Eng. and Min. J., v. 183, July 
 
 1982, pp. 54-58. 
 
 13. Burger, J. (ed.). Mercur is Get- 
 ty's First Gold Mine. Eng, and Min. J., 
 V. 184, No. 10, 1983, pp. 48-51. 
 
 14. Schreiber, H. W, , and M. E. Emer- 
 son. North American Hardrock Gold Depos- 
 its. Eng. and Min. J., v. 185, No. 10, 
 1984, pp. 50-57. 
 
 15. Todd, J. C. New Mines Fuel Aus- 
 tralian Gold Boom, Eng. and Min. J. , v. 
 185, No. 5, 1985, pp. 11-13. 
 
 16. Danne, R. (MINTEK, Johannesburg, 
 Republic of South Africa) . Private com- 
 munication, 1985; available from S. D. 
 Hill, Bureau of Mines, Salt Lake City, 
 UT. 
 
 17. Potter, G. M. , R. S. Shoemaker, 
 K. B. Hall, and D. M. Duncan. Carbon-in- 
 Pulp Processing of Gold and Silver Ores. 
 Min. Eng. (Littleton, CO), pt. 1, v. 33, 
 No. 9, 1981, pp. 1331-1335; pt. 2, v. 33, 
 No. 10, 1981, pp. 1441-1444. 
 
 18. Fleming, C. A. Recent Develop- 
 ments in Carbon-in-Pulp Technology. Pa- 
 per in Hydrometallurgical Research, De- 
 velopment, and Plant Practices, ed. by 
 K. Osseo-Asare and J. D. Miller. Metall. 
 Soc. AIME, 1983, pp. 839-857. 
 
 19. Hussey, S. J., H. B. Salisbury, 
 and G. M. Potter. Carbon-in-Pulp Silver 
 Adsorption From Cyanide Leach Slurries 
 of a Silver Ore. BuMines RI 8268, 1978, 
 22 pp. 
 
 20. . Carbon-in-Pulp Gold Ad- 
 sorption From Cyanide Leach Slurries. 
 BuMines RI 8368, 1979, 22 pp. 
 
 21. Newrick, G. M. , G. Woodhouse, and 
 D. M. G. Dods. Carbon-in-Pulp Versus 
 Carbon-in-Leach. World Min., v. 36, June 
 
 1983, pp. 48-51. 
 
44 
 
 PRECIOUS METALS RECOVERY FROM ELECTRONIC SCRAP AND SOLDER 
 USED IN ELECTRONICS MANUFACTURE 
 
 By B. W, Dunning, Jr. "I 
 
 ABSTRACT 
 
 Electronic scrap from obsolete and/or 
 damaged avionics or from manufacturing 
 sources poses a problem for the owner 
 or generator of the scrap in terms of 
 its fair value. The complexity of this 
 material, as well as the low concentra- 
 tion of precious metals (generally less 
 than 1 pet) , makes it difficult to obtain 
 a representative sample for assay. The 
 owner or generator is therefore dependent 
 on the reliability and competence of a 
 toll refiner to obtain a fair value. 
 
 The Bureau of Mines has investigated var- 
 ious procedures for either concentrating 
 precious metals from electronic scrap in- 
 to an easily assayable form or for re- 
 covering fairly pure gold, silver, or 
 platinum-group metals (PGM). Procedures 
 the Bureau has studied are described 
 and discussed in this paper, including 
 hand dismantling, mechanical processing, 
 pyrometallurgy , hydrometallurgy , and 
 electrometallurgy. 
 
 INTRODUCTION 
 
 Gold, silver, and PGM are widely used 
 in electronic and electrical components 
 to provide long-term reliability. Fabri- 
 cation of equipment used by the military 
 consumes the largest portion of precious 
 metals used in the electronics and elec- 
 trical industry. The disposal of obso- 
 lete and/or damaged military electronics 
 (reportedly totaling more than 10,000 
 st/yr) at a fair value is a pressing 
 problem for the Department of Defense 
 (DOD) . This is partly due to the highly 
 variable and complex nature of military 
 electronic scrap, which makes it diffi- 
 cult to obtain a homogeneous sample for 
 precious metals analysis. Private indus- 
 try has the same problems, but faces few- 
 er constraints in dealing with them. 
 
 The Bureau of Mines, in cooperation 
 with DOD, the National Association of Re- 
 cycling Industries (NARI) , and others. 
 
 has developed many procedures for deter- 
 mining precious metals content in scrap. 
 However, some of these procedures are not 
 economical except under special circum- 
 stances. Initial research to determine 
 precious metals content in electronic 
 scrap relied on hand dismantling and 
 object recognition. Subsequent studies 
 investigated mechanical processing as a 
 method for obtaining metal concentrates 
 containing the major portion of the pre- 
 cious metals. In later studies, these 
 precious metal concentrates were treat- 
 ed using various procedures, includ- 
 ing pyrometallurgy, hydrometallurgy, and 
 electrometallurgy, to further concentrate 
 or recover the precious metals. These 
 procedures are discussed in the following 
 sections, and data concerning the effec- 
 tiveness of each procedure are presented. 
 
 HAND CHARACTERIZATION OF SELECTED AVIONIC MODULES (1-2)2 
 
 In 1977, the Defense Property Disposal 
 Service (of the DOD) initiated a test re- 
 covery program that included having the 
 
 'Supervisory metallurgist, Avondale Re- 
 search Center, Bureau of Mines, Avondale, 
 MD. 
 
 Bureau of Mines hand-process a controlled 
 sample lot containing a variety of "black 
 boxes" (surplus electronic units so 
 
 ^Underlined numbers in parentheses re- 
 fer to items in the list of references at 
 the end of this paper. 
 
45 
 
 designated because of their black out- 
 er cases). The program was to include 
 determination of base and precious met- 
 als content, removal of salable modules, 
 and a monetary evaluation of each box. 
 Several identical lots were prepared 
 with components intact to assure a basis 
 for direct comparison. The Bureau re- 
 ceived one lot of boxes for hand dis- 
 mantling to determine the potential yield 
 of base and precious metals from each 
 unit. An identical lot of boxes was 
 placed on display at the Defense Con- 
 struction Supply Center, Columbus, OH, 
 for private industry to bid on determina- 
 tion of the following: 
 
 1. Relative cost-effectiveness of and 
 estimated cost per item for currently 
 practicable methods for precious metals 
 segregation, identification, and recovery 
 and for serviceability testing of usable 
 modules from individual boxes, 
 
 2. Reutilization and sales potential 
 of usable modules and components, with 
 
 sales potential supported by quotations 
 from prospective purchasers. 
 
 All interested industry groups indi- 
 cated that the sample lot was too low 
 in value to warrant bidding and too 
 old for reutilization of usable mod- 
 ules. More than half of the sample con- 
 sisted of radio receivers, transmitters, 
 tuners, and power supplies; the remain- 
 der consisted of miscellaneous naviga- 
 tional and communication equipment. All 
 units appeared to have been produced 
 prior to 1957 and did not contain any 
 printed circuits. The weights for in- 
 dividual units in the sample lot ranged 
 from 2-1/2 to 58 lb. The total weight 
 of the sample lot was 726 lb. The avi- 
 onlc units came from the military air- 
 craft storage located in the Arizona 
 desert near Tucson. A summary of the 
 materials composition of the 36 hand- 
 dismantled avionic units is listed in 
 table 1. 
 
 MECHANICAL PROCESSING OF ELECTRONIC SCRAP AND GOLD AND SILVER 
 DISTRIBUTION IN THE VARIOUS FRACTIONS (3-5) 
 
 Approximately 5 st of general avionic 
 scrap "black boxes" was mechanically pro- 
 cessed through a series of unit opera- 
 tions. These unit operations included a 
 hammer mill, air classif ier-baghouse, 
 magnetic separator-trommel, vibrating 
 screen, rolls crusher, wire separator 
 screen, magnetic precleaner, eddy-current 
 separator, and high-tension separator. 
 The various materials recovered from 
 these unit operations and their distribu- 
 tion are listed in table 2, along with 
 the assay and the amount of gold and sil- 
 ver in each material fraction. Three ma- 
 terial fractions, the lights, wire bun- 
 dles, and the metallics from high-tension 
 separation, contained most of the gold 
 and silver. It was determined that these 
 materials, representing 34 pet of the 
 original 5 st, could be economically pro- 
 cessed by a toll refiner to obtain credit 
 for the precious metals. 
 
 Hand-segregated printed cards and elec- 
 trical plugs and connectors were also 
 processed through a modified series of 
 mechanical unit operations. Rolls crush- 
 ing and magnetic precleaning were not 
 deemed necessary for processing these 
 items. The amount and distribution of 
 materials recovered from circuit cards 
 using the various unit operations are 
 shown in table 3, along with the assay 
 and the amount of gold and silver in each 
 material fraction; the results shown are 
 from processing approximately 920 lb of 
 printed-circuit cards. 
 
 The amount and distribution of materi- 
 als recovered from mechanically process- 
 ing some 600 lb of electrical plugs and 
 connectors are listed in table 4, along 
 with the assay and the amount of gold and 
 silver in these components. Most of the 
 gold and silver was found in the high- 
 tension metal and baghouse lights. 
 
46 
 
 TABLE 1. - Summary of materials composition of hand-dismantled avionics units, 
 weight percent 
 
 Uniti 
 
 Aluminum 
 base 
 
 Copper 
 base 
 
 Magnetic 
 metals 
 
 Stainless 
 steel 
 
 Nonmetals 
 
 Fraction con- 
 taining pre- 
 cious metals^ 
 
 Receiver-transmitter 
 
 Tuner, radio 
 
 Tuner, radio 
 
 Tuner, radio 
 
 Radio receiver 
 
 Converter 
 
 Keyer 
 
 Amplifier 
 
 Video amplifier 
 
 Video decoder 
 
 Radio receiver 
 
 Receiver-transmitter 
 Control transmitter. 
 
 Video coder 
 
 Indicator 
 
 Tuner, radio 
 
 Tuner, radio 
 
 Tuner, radio 
 
 Tuner, radio 
 
 Receiver 
 
 Coder transmitter 
 
 set 
 
 Receiver-transmitter 
 Azimuth indicator, . . 
 
 Receiver 
 
 Receiver-transmitter 
 
 Indicator 
 
 Azimuth indicator. . . 
 
 Power supply 
 
 Receiver-transmitter 
 
 Power supply 
 
 Power supply 
 
 Power supply 
 
 Storage unit 
 
 N^ compass 
 
 Inverter 
 
 Electron tube 
 
 20. 
 57. 
 56. 
 54. 
 25. 
 47. 
 28. 
 32. 
 29. 
 51. 
 28. 
 56. 
 32. 
 54. 
 39. 
 59. 
 62. 
 61. 
 51. 
 34. 
 
 45. 
 36. 
 36. 
 25. 
 32. 
 43. 
 40. 
 27. 
 28. 
 6. 
 28. 
 25. 
 58. 
 16. 
 32. 
 14. 
 
 35.8 
 20.8 
 20.1 
 17.6 
 28.3 
 23.5 
 23.6 
 22.2 
 23.5 
 13.3 
 25.9 
 17.0 
 22.4 
 16.0 
 12.8 
 16.5 
 12.4 
 14.7 
 17.7 
 17.4 
 
 15.3 
 18.8 
 19.8 
 21.7 
 17.7 
 17.0 
 17.5 
 19.0 
 18.1 
 24.4 
 16.8 
 15.9 
 14.1 
 
 5.1 
 19.4 
 
 8.4 
 
 19.1 
 
 9.5 
 
 9.1 
 
 10.5 
 
 29.0 
 
 11.0 
 
 21.0 
 
 23.8 
 
 19.3 
 
 11.5 
 
 23.5 
 
 13.5 
 
 12.9 
 
 9.0 
 
 12.9 
 
 8.3 
 
 8.3 
 
 9.3 
 
 14.1 
 
 26.7 
 
 22.5 
 23.1 
 21.1 
 30.1 
 21.2 
 19.6 
 20.8 
 31.1 
 28.6 
 47.6 
 34.5 
 33.3 
 15.7 
 43.8 
 41.4 
 69.4 
 
 4.3 
 4.4 
 5.3 
 7.5 
 2.3 
 3.2 
 5.7 
 2.3 
 8.9 
 8.5 
 1.5 
 1.3 
 13.7 
 3.4 
 3.9 
 5.5 
 8.5 
 5.9 
 7.9 
 1.7 
 
 1.4 
 3.3 
 4.4 
 3.4 
 3.4 
 3.0 
 4.3 
 2.3 
 4.0 
 2.5 
 
 .6 
 3.7 
 
 .5 
 33.3 
 
 .8 
 5.5 
 
 20.3 
 
 7.9 
 
 8.6 
 
 9.5 
 
 15.3 
 
 14.8 
 
 21.2 
 
 18.7 
 
 18.9 
 
 15.1 
 
 20.4 
 
 11.5 
 
 18.9 
 
 17.3 
 
 30.9 
 
 10.1 
 
 8.8 
 
 8.3 
 
 9.1 
 
 19.6 
 
 15.5 
 18.3 
 18.5 
 18.9 
 25.7 
 16.5 
 17.1 
 19.7 
 20.5 
 19.2 
 19.4 
 21.9 
 11.3 
 1.7 
 5.8 
 2.3 
 
 14.5 
 
 13.5 
 
 13.4 
 
 11.6 
 
 11.1 
 
 8.2 
 
 5.2 
 
 5.2 
 
 4.7 
 
 3.6 
 
 3.5 
 
 2.8 
 
 2.8 
 
 2.8 
 
 2.6 
 
 2.5 
 
 2.3 
 
 2.3 
 
 2.2 
 
 2.2 
 
 2.2 
 
 1.9 
 
 1.8 
 
 1.8 
 
 1.5 
 
 1.4 
 
 1.3 
 
 1.3 
 
 1.2 
 
 1.1 
 
 1.0 
 
 .8 
 
 .3 
 
 .2 
 
 .2 
 
 
 
 'where units of the same kinc 
 individual unit. 
 
 ^Percentage of the total 
 hand segregation of silver- 
 
 are listed more than once, each listing represents an 
 
 black box we 
 and gold-co 
 
 ight that contained precious metals, based on 
 ated components through visual examination. 
 
47 
 
 TABLE 2. - Material distribution, concentration, and weight of contained precious 
 metals from fractions of mechanically beneficiated general avionic scrap 
 
 Fraction 
 
 Weight, 
 lb 
 
 Distribution, 
 wt pet 
 
 Concentration, 
 wt pet 
 
 Contained precious 
 metals, tr oz 
 
 
 Au 
 
 Ag 
 
 Au 
 
 Ag 
 
 Baghouse lights 
 
 Wire bundles ............. 
 
 930 
 860 
 
 1,204 
 1,825 
 
 425 
 
 1,161 
 
 965 
 
 552 
 444 
 773 
 
 1,318 
 
 23 
 
 8.9 
 8.2 
 
 11.5 
 17.4 
 
 4.0 
 
 11.1 
 
 9.2 
 
 5.3 
 
 4.2 
 7.4 
 
 12.6 
 
 .2 
 
 0.021 
 .021 
 
 .015 
 0) 
 
 .064 
 
 ND 
 
 .003 
 
 .099 
 .074 
 .031 
 
 .017 
 
 0.33 
 .77 
 
 .07 
 (M 
 
 .41 
 
 .017 
 
 .64 
 
 2.01 
 1.96 
 1.09 
 
 .30 
 
 2.8 
 2.6 
 
 2.6 
 0) 
 
 4.0 
 ND 
 .42 
 
 8.0 
 4.8 
 3.5 
 
 3.3 
 
 44.7 
 96.6 
 
 Magnetics: 
 
 Minus 1/2 in 
 
 12.3 
 
 Minus 1 plus 1/2 in.... 
 Eddy-current: 2 
 
 Precleaner magnetics... 
 Aluminum. .••.....•..••. 
 
 0) 
 
 25.4 
 2.9 
 
 Middles 
 
 90.1 
 
 High-tension metal: ^ 
 
 Minus 1/4 in. .......... 
 
 161.7 
 
 Minus 1/2 plus 1/4 in.. 
 Minus 1 plus 1/2 in.... 
 High-tension rejects 
 
 (minus 1 in) ^ 
 
 Hammer-mill knockout box 
 contents 
 
 127.0 
 122.9 
 
 57.7 
 
 (M 
 
 Composite, total or 
 average 
 
 10,480 
 
 100.0 
 
 .021 
 
 .48 
 
 32.02 
 
 741.3 
 
 ND Not detected. 
 
 'visual observation of this fraction showed no evidence of precious metals except 
 for a few large power transistors. 
 
 ^Fractions obtained using ramp-type eddy-current separator. 
 
 ^Fractions obtained using high-tension separator. 
 
 ^Not determined. 
 
 TABLE 3. - Material distribution, concentration and weight of contained precious 
 metals in fractions of mechanically beneficiated printed-circuit cards 
 
 Fraction 
 
 Weight, 
 lb 
 
 Distribution, 
 wt pet 
 
 Concentration, 
 wt pet 
 
 Contained precious 
 metals, tr oz 
 
 
 Au 
 
 Ag 
 
 Au 
 
 Ag 
 
 Baghouse lights . . . . < 
 
 Wire bundles 
 
 214.6 
 
 11.2 
 
 283.6 
 
 16.6 
 
 324.9 
 
 72.0 
 
 23.2 
 
 1.2 
 
 30.7 
 
 1.8 
 
 35.2 
 
 7.8 
 
 0.075 
 .068 
 .030 
 
 ND 
 
 .150 
 
 .023 
 
 0.61 
 .97 
 .05 
 
 .01 
 
 1.55 
 
 .37 
 
 2.35 
 
 .11 
 
 1.24 
 
 ND 
 
 7.11 
 
 .24 
 
 19.1 
 1.6 
 
 Magnetics (minus 1/2 in). 
 Eddy-current aluminum 
 
 (minus 1/2 in) 
 
 High-tension metal (minus 
 
 1/2 in) 
 
 2.1 
 .02 
 73.4 
 
 High-tension rejects 
 (minus 1/2 in) 
 
 3.9 
 
 Composite, total or 
 average 
 
 922.9 
 
 99.9 
 
 .082 
 
 .74 
 
 11.05 
 
 100.1 
 
 ND Not detected. 
 
48 
 
 TABLE 4. - Material distribution, concentration, and weight of contained precious 
 metals in fractions of mechanically beneficiated electrical plugs and connectors 
 
 Fraction 
 
 Weight, 
 lb 
 
 Distribution, 
 wt pet 
 
 Concentration, 
 wt pet 
 
 Contained precious 
 metals, tr oz 
 
 
 Au 
 
 Ag 
 
 Au 
 
 Ag 
 
 Baghouse lights 
 
 Wire bundles 
 
 81.6 
 
 6.6 
 
 53.2 
 
 66.6 
 
 199.8 
 
 191.8 
 
 13.6 
 1.1 
 8.9 
 
 11.1 
 
 33.3 
 
 32.0 
 
 0.081 
 .043 
 .020 
 
 ND 
 
 .137 
 
 ND 
 
 0.88 
 
 1.34 
 
 .10 
 
 .015 
 1.12 
 ND 
 
 0.96 
 .04 
 .16 
 
 ND 
 
 3.99 
 
 ND 
 
 10.5 
 1.3 
 
 Magnetics (minus 1/2 in). 
 Eddy-current aluminum 
 
 (minus 1/2 in) 
 
 High-tension metal (minus 
 
 1/2 in) 
 
 High-tension rejects 
 
 (minus 1/2 in)' 
 
 .78 
 
 .15 
 32.6 
 
 ND 
 
 Composite 
 
 599.6 
 
 100.0 
 
 .059 
 
 .52 
 
 5.15 
 
 45.3 
 
 ND Not detected. 
 
 'visual observation of the high-tension rejects (minus 1/2 in) indicated that the 
 shattered pieces of plastic, ceramic, hard rubber, and other types of insulator mate- 
 rial contained no precious-metals-bearing contact pins after shredding. 
 
 In processing either whole avionic 
 units or hand-segregated circuit cards , 
 electrical plugs, and connectors, the 
 greatest amount of gold and silver is 
 always concentrated in the high-tension 
 metal product. Most of the remaining 
 gold and silver concentrates in the wire 
 and baghouse lights. These fractions are 
 best handled by a toll refiner. A 
 toll refiner will first incinerate these 
 
 fractions separately to remove the or- 
 ganics. After incineration, the high- 
 tension metal product and wire are dis- 
 solved in a molten heel of copper. The 
 melt is then thoroughly mixed and cast 
 into ingots for assay of the precious 
 metals. The baghouse lights are also in- 
 cinerated, then ball or rod milled; the 
 resulting powder is then blended, coned, 
 quartered, and assayed. 
 
 HYDROMETALLURGICAL RECOVERY FROM ELECTRONIC SCRAP 
 
 GOLD AND SILVER FROM HIGH-TENSION 
 METALLIC FRACTION (6-7) 
 
 three stages with 20-vol-pct H2 SO4 to re- 
 move the copper. 
 
 The high-tension metallics from me- 
 chanically processed avionic units, rep- 
 resenting about 17 pet of the units' 
 total weight, contained slightly more 
 than 50 pet of the total gold and sil- 
 ver. The high-tension fraction, with 
 its high surface-area-to-weight ratio, is 
 ideal for upgrading by leaching. The 
 elemental composition of this fraction is 
 listed in table 5. The major base metals 
 in the high-tension fraction were copper, 
 aluminum, and iron, in that order. Alu- 
 minum is removed by leaching with 20- 
 wt-pct NaOH. After washing, the residue 
 is countercurrently pressure leached in 
 
 TABLE 5. - Elemental composition of 
 high-tension metallic fraction 
 from avionic scrap 
 
 Ag. 
 Al. 
 Au. 
 Cr, 
 Cu. 
 
 Cone, 
 
 wt pet 
 
 1.37 
 27.2 
 .12 
 .1 
 38.4 
 
 Fe. 
 Ni. 
 Pb. 
 Sn. 
 
 Cone, 
 wt pet 
 
 9.4 
 
 3.2 
 
 .2 
 
 2.1 
 
 NOTE. — Remainder was mostly Si as Si02 
 (fiberglass and plastic filler). 
 
49 
 
 The distribution of metal values in the 
 leaeh solutions and in the residues is 
 listed in table 6. Copper is reeovered 
 from the pregnant leaeh liquor by eemen- 
 tation with transformer steel from the 
 coarse magnetic fraction of mechanically 
 processed avionics. The cement copper 
 is a good grade and assays better than 
 90 pet. 
 
 TABLE 6. - Concentration and distribu- 
 tion of metal values in leaeh solu- 
 tions and residues 
 
 
 Leach solutions 
 
 Residues 
 
 Elements 
 
 Cone, 
 
 Distri- 
 
 Cone, 
 
 Distri- 
 
 
 g/L 
 
 bution, 
 pet 
 
 wt pet 
 
 bution, 
 pet 
 
 Ag 
 
 <0.001 
 
 <0.004 
 
 8.13 
 
 99.1 
 
 Au 
 
 ND 
 
 ND 
 
 .74 
 
 >99.9 
 
 Cr 
 
 .0004 
 
 6.2 
 
 .42 
 
 94.0 
 
 Cu 
 
 36.3 
 
 89.3 
 
 24.4 
 
 10.9 
 
 Fe 
 
 2.9 
 
 29.0 
 
 39.6 
 
 71.0 
 
 Ni 
 
 1.02 
 
 38.8 
 
 9.0 
 
 61.1 
 
 Pb 
 
 .003 
 
 1.4 
 
 1.1 
 
 98.6 
 
 Pd 
 
 ND 
 
 ND 
 
 .21 
 
 99.9 
 
 Sn 
 
 .14 
 
 7.9 
 
 9.5 
 
 91.1 
 
 ND Not detected. 
 
 Silver is reeovered from the pressure- 
 leached residue by dissolution in 50- 
 vol-pet HNO3 and subsequent precipitation 
 with NaCl as AgCl. This AgCl is reduced 
 to silver metal by mixing with Na2C03 and 
 heating to 600° C. Copper not extracted 
 during H2SO4 leaching is reeovered from 
 the silver-free HNO3 leaeh solution by 
 cementation with steel scrap. Gold is 
 extracted from the HNO3 leaching residue 
 with aqua regia and precipitated with 
 NaHS03. Analyses of the gold and silver 
 products and the final residue is listed 
 in table 7. The final residue consisted 
 mostly of acid-insoluble noncombustibles 
 (mostly silica) , stainless steel, and 
 tin. 
 
 PLATINUM-GROUP METALS (_8) 
 
 Hydrometallurgical techniques have al- 
 so been investigated as a means for 
 recovering platinum-group metals from 
 electronic scrap. 
 
 Telephone Relay Scrap 
 
 NARI estimates that about 1,000 st/yr 
 of telephone relay scrap is available for 
 processing. Palladium, gold, and plati- 
 num are found in the contact points, 
 which are brazed to cupronickel wires. 
 These contact points represent approxi- 
 mately 0.06 pet of the modular weight. 
 Mechanical processing, including shred- 
 ding, air classification, magnetic sep- 
 aration, screening, and high-tension sep- 
 aration (HTS), produced an HTS metallic 
 fraction containing most of the relay 
 contact points. A portion of this HTS 
 concentrate representing 8.7 pet of the 
 original scrap was pressure leached in 
 stainless steel autoclaves in a two-stage 
 countercurrent arrangement with solid- 
 liquid separation between each stage. 
 The purpose of the first-stage leaeh was 
 to react fresh HTS concentrate with the 
 second-stage liquor to produce a pregnant 
 liquor with a low concentration of H2 SO4 
 (less than 10 g/L). Partially leached 
 solids from the first-stage leach were 
 fed to the second stage, where they were 
 leached with a fresh H2SO4-HNO3 solution. 
 Measured quantities of HTS-PGM con- 
 centrate (usually 120 g), 1 L of H2 SO4 
 solution (20 vol pet), and HNO3 (20 mL 
 to 120 g concentrate) were charged to 
 the autoclave and heated to 90° C with a 
 lOO-lbf/in^ air overpressure. Air was 
 sparged through the autoclave (about 10 
 bubbles per second) during all leaches. 
 After leaching, the slurry was removed 
 from the autoclave and filtered, and the 
 products were analyzed. Increasing the 
 reaction time from 1 to 4 h increased 
 base metal extraction from 96 to 99 pet. 
 Quantitative analyses of the residue and 
 filtrate are listed in table 8. The res- 
 idue, representing 0.2 pet of the HTS-PGM 
 concentrate, was suitable for shipping to 
 a toll refiner for recovery of silver, 
 gold, and PGM. 
 
 Reed Switches 
 
 NARI sources estimate that approximate- 
 ly 10 St of reed switches, a valuable 
 scrap material, is available for process- 
 ing annually. Reed switches consist of 
 
50 
 
 TABLE 7. - Semiquantitative spectrochemical analyses of products 
 from pressure-leached residue, weight percent 
 
 Element 
 
 Pro 
 
 duct 
 
 Final residue 
 
 
 Gold 
 
 Cement silver^ 
 
 
 Ag 
 
 Al 
 
 0.03 - 0.3 
 .003- .03 
 
 >10 
 .003- .03 
 .003- .03 
 .03 - .3 
 .003- .03 
 .01 - .1 
 
 1 - 10 
 .1 - 1 
 .03 - .3 
 
 >10 
 0.03 - .3 
 .0003- .003 
 
 ND 
 .03 - .3 
 .03 - .3 
 .003 - .3 
 .3-3 
 
 ND 
 .1 - 1 
 .3-3 
 
 0.0003- 0.003 
 .3-3 
 
 Au 
 
 ND 
 
 Cr 
 
 .3-3 
 
 Cu 
 
 .1-1 
 
 Fe 
 
 .3-3 
 
 Ni 
 
 .01 - .1 
 
 Pb 
 
 .1-1 
 
 Pd 
 
 ND 
 
 Si 
 
 >10 
 
 Sn 
 
 1 - 10 
 
 ND Not detected. 
 
 ^Quantitative analysis of remaining cement silver after semi- 
 quantitative spectrochemical analysis, in weight percent: 97.5 
 Ag, 0.028 Al, 0.53 Cu, 0.12 Pb, and 1.82 Sn. 
 
 TABLE 8. - Quantitative analyses of 
 residue and filtrate from leaching 
 telephone relays 
 
 Element 
 
 Concentration 
 
 
 Residue, wt 
 
 pet 
 
 Filtrate, g/L 
 
 Ag 
 
 1.3 
 
 
 0.001 
 
 Au 
 
 14.8 
 
 
 .0004 
 
 Cu 
 
 1.3 
 
 
 67.9 
 
 Cr 
 
 <.003 
 
 
 .0003 
 
 Mn 
 
 ND 
 
 
 .03 
 
 Ni 
 
 .09 
 
 
 14.4 
 
 Pd 
 
 77.7 
 
 
 .0007 
 
 Pt 
 
 .29 
 
 
 .009 
 
 ND Not detected. 
 
 two magnetizable reeds with their extrem- 
 ities fused in a glass envelope. The in- 
 ner ends of the reeds are plated with 
 gold and then rhodium. A simple rolls 
 crushing step followed by magnetic sepa- 
 ration of the reeds from the glass powder 
 was the only preprocessing necessary to 
 prepare the reed switches for hydrometal- 
 lurgical treatment. Assay of the reed 
 switches showed them to be, in weight 
 percent, 49 Co, 48 Fe, 2 V, 0.5 Au, and 
 0.4 Rh. 
 
 The approach to recovering gold and 
 rhodium from this scrap was to dis- 
 solve the cobalt-iron substrate in either 
 HCl (1:1) or HNO3 (1:2) using ultrasonic 
 agitation. Dissolution of the base-metal 
 
 substrate was relatively fast except for 
 the working face of the reed plated with 
 gold and rhodium. Complete removal of 
 base metals required 12 h of continuous 
 leaching. The insoluble residue was a 
 fine gold-rhodium sand assaying approxi- 
 mately 50 pet Au and 50 pet Rh. A spec- 
 trochemical analysis of the gold-rhodium 
 sand is listed in table 9. Gold was 
 readily removed from the Rh with aqua 
 regia, leaving a pure 99.8-pct-Rh sand. 
 Recovery of the gold was accomplished by 
 precipitating with NaHS03 ; however, other 
 methods of recovering the gold are avail- 
 able. The byproduct metals cobalt and 
 vanadium can be recovered from the iron- 
 cobalt-vanadium solution using existing 
 technology. 
 
 TABLE 9. - Semiquantitative spectro- 
 chemical analysis of gold-rhodium 
 sand from reed switches 
 
 Ele- 
 ment 
 
 Cone, 
 wt pet 
 
 Ag... 
 
 . 0.003- 0.03 
 
 Al... 
 
 .003- .03 
 
 Au. . , 
 
 >10 
 
 Co... 
 
 .03 - .3 
 
 Fe... 
 
 .01 - .1 
 
 Mg... 
 
 .001- .01 
 
 Ele- 
 ment 
 
 Mn.. 
 Pb.. 
 Rh.. 
 Si.. 
 Sn. . 
 
 Cone, 
 wt pet 
 
 0.001- 0.01 
 .003- .03 
 
 >10 
 .01 - .1 
 .03 - .3 
 
51 
 
 COPPER CEMENTATION WITH SELECTED MATERIALS FROM AV IONIC SCRAP (9) 
 
 A process using either brittle aluminum 
 base or magnetic metallics from elec- 
 tronic scrap as a copper precipitant in 
 acidulated CUSO4 solution was developed 
 to separate and then upgrade a high-grade 
 cement copper containing all or most of 
 the precious metals. Aluminum-bearing 
 obsolete avionic assemblies and plugs and 
 connectors removed from electronic scrap 
 were incinerated and then melted and cast 
 to form brittle ingots. An assay of 
 these incinerated and melted items is 
 listed in table 10. The brittle ingots 
 were broken up and rolls-crushed to minus 
 35 mesh. Copper cementation took place 
 when this material was agitated in the 
 acidulated CUSO4 solution in a counter- 
 current system. The precipitate con- 
 tained better than 90 pet Cu and all of 
 the associated precious metals. The ce- 
 ment copper thus obtained can be bene- 
 ficiated by fire refining to produce 
 high-grade anode material for subsequent 
 electrorefining to cathode copper. An 
 anode mud is produced from which the pre- 
 cious metals can be recovered using stan- 
 dard procedures. 
 
 TABLE 10. - Analyses of ingots produced 
 by Incineration and melting of mili- 
 tary electronic scrap 
 
 Analysis, tr oz/st: 
 
 Ag 
 
 Au 
 
 Concentration, wt pet 
 
 Al 
 
 Cu 
 
 Fe 
 
 Mn 
 
 Ni 
 
 Zn 
 
 Insol 
 
 ^Not determined. 
 
 The magnetic fraction of shredded elec- 
 tronic scrap, which is mostly thin sili- 
 con steel laminae from transformers, is 
 
 Plugs and 
 connec- 
 tors 
 
 Avionic 
 assem- 
 blies 
 
 157.0 
 0.16 
 
 38.4 
 
 24.0 
 
 26.3 
 
 (') 
 
 0) 
 
 0.23 
 
 4.7 
 
 an excellent copper cementation agent 
 when it is agitated in an acidulated 
 CUSO4 solution. It also contains nickel- 
 alloy transistor caps that contain some 
 gold and silver. A representative sample 
 of the magnetic fraction from shredded 
 electronic scrap was melted, cast, and 
 sampled for assay. An assay of this ma- 
 terial is listed in table 11. 
 
 TABLE 11, - Analysis of a representa- 
 tive sample of shredded magnetic 
 scrap 
 
 Value 
 
 Analysis, tr oz/st: 
 
 Ag 7,2 
 
 Au 5,96 
 
 Concentration, wt pet: 
 
 Cr 0,7 
 
 Cu 0,7 
 
 Fe 75,8 
 
 Ni 11,5 
 
 Insol 4,4 
 
 Copper cementation was most efficient 
 when shredded magnetic scrap was used in 
 a tumbler-type system, since a vigorous 
 scrubbing action was needed to maintain 
 the precipitation reaction. An analysis 
 of the combined precipitate showed it to 
 contain, in weight percent, 89 Cu, 1,1 
 Fe, 0,32 Pb, 0,36 Sn, 1,5 insolubles, and 
 in ounces per short ton, 4,0 Au and 11,5 
 Ag. 
 
 The total precipitate was melted in an 
 induction furnace without flux. An an- 
 ode, representing 90 pet of the charge 
 weight, was obtained from the melt; the 
 balance, containing most of the impuri- 
 ties, remained as a sinter. The anode 
 was placed in a polypropylene sock and 
 electroref ined in a solution containing 
 150 g/L H2SO4 and 40 g/L Cu at a current 
 density of 12 A/f t^ , Previous electro- 
 refining tests using these conditions 
 produced a dense, smooth cathode deposit. 
 Assays of the various products are listed 
 in table 12, The anode mud can be fur- 
 ther refined using standard procedures 
 for gold and silver recovery. 
 
52 
 
 TABLE 12. - Analyses of electro- 
 refining products 
 
 
 Anode 
 
 Cathode 1 
 
 Anode 
 mud 
 
 Analysis, 
 
 Ag 
 
 Au 
 
 tr oz/st: 
 
 • ••••••••• 
 
 12.1 
 4.3 
 
 99.4 
 
 0.13 
 
 0.104 
 
 0.047 
 
 0.05 
 
 ND 
 ND 
 
 99.9 
 
 ND 
 ND 
 ND 
 ND 
 
 1,140 
 412 
 
 Analysis, 
 Cu 
 
 pet: 
 
 57.0 
 
 Fe 
 
 ND 
 
 Pb 
 
 Sn 
 
 0.7 
 0.316 
 
 Insol 
 
 NS 
 
 TABLE 13. - Analyses of products ob- 
 tained from melting and electrolyti- 
 cally solubilizing unreacted magnetic 
 scrap residue 
 
 ND Not determined. 
 
 'a spectrographic analysis of the cath- 
 ode copper showed the following impuri- 
 ties (estimated): 0.001 wt pet Mg and 
 0.01 wt pet Ca. 
 
 The cleaned, unreacted residue from the 
 copper cementation with shredded magnetic 
 scrap (amounting to 11.0 pet of the orig- 
 inal charge) was melted in an induction 
 furnace and cast into an anode. An assay 
 of the metal is listed in table 13. 
 
 The anode was suspended in a elec- 
 trolytic cell fitted with a graphite 
 
 Analysis, tr oz/st: 
 
 Ag 
 
 Au 
 
 Analysis, wt pet: 
 
 Co 
 
 Cr 
 
 Cu 
 
 Fe 
 
 Ni 
 
 Insol 
 
 Resi- 
 due^ 
 
 35 
 12 
 
 0.7 
 
 0.5 
 
 11.0 
 
 40.0 
 
 46.4 
 
 Na 
 
 Anode 
 mud 
 
 239 
 129 
 
 NA 
 
 NA 
 
 55.0 
 
 17.0 
 
 18.0 
 
 8.0 
 
 Solution, 
 g/L 
 
 NA 
 NA 
 
 1.9 
 
 0.9 
 
 0.05 
 
 50 
 
 61 
 
 NA 
 
 NA Not analyzed. 
 
 ^Cleaned, unreacted scrap residue. 
 
 cathode and partially solubilized in an 
 electrolyte containing 150 g/L H2SO4. 
 The metals in solution (table 13) could 
 be treated by presently used technology 
 to recover the nickel, cobalt, and chro- 
 mium. The anode mud (table 13) can be 
 treated using standard procedures for the 
 recovery of the precious metals. 
 
 SWEATING AVIONIC SCRAP TO PRODUCE ALUMINUM BULLION 
 FOR FUSED SALT ELECTROLYSIS (10) 
 
 The Bureau developed and tested two 
 molten-salt electrorefining procedures 
 for processing aluminxom ingots sweated 
 from avionic scrap in order to recover a 
 high-quality aluminum and concentrate the 
 gold and silver in the aluminum-depleted 
 anodes. An analysis of one of the ingots 
 is listed in table 14. 
 
 One system used a three-layer cell. 
 The three molten layers were separated 
 because of differences in the densities 
 of the molten anode material, the molten 
 salt electrolyte, and the molten refined 
 aluminum. At the operating temperature 
 range of 750° to 850° C, the approximate 
 density was 3.3 g/cm-' for the molten 
 electronic scrap, 2.7 g/cva? for the mol- 
 ten electrolyte, and 2.3 g/cm^ for the 
 molten refined aluminum. The electrolyte 
 contained, in weight percent, 60 BaCl2, 
 17 NaF, and 23 AIF3. The recovered alu- 
 minum had a purity of 99.8 pet. The com- 
 position of the anode residue from the 
 three-layer cell (representing 33 pet of 
 
 the electronic scrap ingot) is listed in 
 table 15. The anode residue can be read- 
 ily refined to copper bullion suitable 
 for further treatment by aqueous acid 
 electrolysis to separate precious metals 
 and copper. 
 
 TABLE 14. - Composition of electronic 
 scrap ingot 
 
 Value 
 
 Analysis, tr oz/st: 
 
 Ag 119 
 
 Au 12 
 
 Concentration, wt pet: 
 
 Al 69.69 
 
 Cu 19.98 
 
 Fe 0.50 
 
 Mg 0.13 
 
 Mn 0.20 
 
 Ni 0.22 
 
 Pb 1.07 
 
 Si 2.57 
 
 Sn 1.19 
 
 Zn 4.02 
 
53 
 
 TABLE 15. - Composition of total anode 
 residue, three-layer cell 
 
 Value 
 
 Analysis, tr oz/st: 
 
 Ag 378 
 
 Au 37 
 
 Concentration, wt pet: 
 
 Al 6.93 
 
 Cu 61.61 
 
 Fe 2.03 
 
 Mg 0.007 
 
 Mn 0.32 
 
 Ni 0.61 
 
 Pb 1.88 
 
 Si 6.48 
 
 Sn 6.75 
 
 Zn 10. 16 
 
 A second system used a compartmented 
 cell that provided separate compartments 
 for the anode and cathode metals instead 
 of the density separation used in the 
 three-layer cell. The cell was operated 
 in the range of 750° to 800° C. The 
 electrolyte consisted of an equimolar 
 mixture of NaCl and KCl, with enough 
 AICI3 to provide a 1- to 2-wt-pct Al 
 concentration in the electrolyte. The 
 
 recovered aluminum had a purity of 99.6 
 pet. The composition of the anode resi- 
 due from the compartmental cell (repre- 
 senting 32 pet of the electronic scrap 
 ingot) is listed in table 16. This anode 
 residue can also be refined to copper 
 bullion suitable for electroref ining. 
 The two fused-salt electroref ining pro- 
 cesses are technically feasible; however, 
 economic evaluations have not been made. 
 
 TABLE 16. - Composition of total anode 
 residue, compartmental cell 
 
 Value 
 
 Analysis, tr oz/st: 
 
 Ag 365 
 
 Au 35 
 
 Concentration, wt pet: 
 
 Al 12.2 
 
 Cu 60.9 
 
 Fe 1.66 
 
 Mg <0.01 
 
 Mn 0.45 
 
 Ni 0.68 
 
 Pb 3.27 
 
 Si 8.04 
 
 Sn 3.64 
 
 Zn 7.77 
 
 INCINERATION, CAUSTIC LEACHING, SMELTING, AND 
 ELECTROREFINING OF AVIONIC SCRAP (11) 
 
 From a metallurgical standpoint, avi- 
 onic scrap is a complex mixture of vari- 
 ous metals, mostly copper, aluminum, and 
 iron, attached to, covered with, or mixed 
 with diverse types of plastics and ceram- 
 ics. Precious metals occur as platings 
 of various thicknesses, in relay contact 
 points, on switch contacts and wires, and 
 in solders. 
 
 The plastics and organics can be elimi- 
 nated prior to smelting by incinera- 
 tion at 400° to 500° C in a gas-fired 
 furnace. The furnace must be equipped 
 with an afterburner and a scrubber in 
 order to meet antipollution regulations. 
 Iron alloys are removed with a drum 
 magnet. The aluminum content, generally 
 about 35 to 40 pet in avionics, may be 
 removed by caustic leaching with regener- 
 ation of the caustic. Leaching is not 
 necessary unless there is a ready market 
 for the AI2O3. Smelting of the inciner- 
 ated scrap, with or without the aluminiim 
 
 removed, is done to produce a homogeneous 
 assayable product that can be sold to a 
 custom smelter. An ingot produced from 
 electronic scrap, without the aluminum 
 removed, assayed 30 wt pet Cu, 18 wt pet 
 Fe, and 32 wt pet Al, and 120 tr oz/st 
 Ag and 1 tr oz/st Au. Smelting another 
 batch of electronic scrap low in alumi- 
 num produced an ingot that assayed 85 wt 
 pet Cu, 4 wt pet Fe, and 0.2 wt pet Al, 
 and 333 tr oz/st Ag and 26 tr oz/st Au. 
 These two assays illustrate the hetero- 
 geneity of electronic scrap composition. 
 The remelted ingots, with or without the 
 aluminum removed, can be further refined 
 by slag additions and blowing air through 
 the melt. The resulting copper bullion 
 is refined electrolytically , and the an- 
 ode slimes, which are rich in precious 
 metals, are sent to a toll refiner for 
 final processing. 
 
 Economic evaluation of the various 
 procedures for processing avionic scrap 
 
 B lWjWIgWI IIHHg 
 
54 
 
 indicates that incineration followed by 
 smelting to form an assayable ingot 
 and selling this product to a custom 
 
 electroref iner is the most cost-effective 
 procedure. 
 
 TREATMENT OF SPENT TIN-LEAD SOLDER FROM MANUFACTURE OF ELECTRONIC 
 PRINTED CIRCUIT CARDS TO RECOVER GOLD (12-13) 
 
 Estimates indicate that 2,000 st of 
 spent solder containing about 120,000 
 tr oz Au is generated annually (12). 
 One method for recycling the tin-led 
 solder and recovering the gold is fused- 
 salt electrolysis (12). Electroref ining 
 of the spent solder is carried out at 
 450° to 500° C with a molten KCl-SnCl2- 
 PbCl2 electrolyte. The spent solder is 
 charged to the anode container. Both 
 anode and cathode electrodes are tung- 
 sten. Electrolyte composition is, in 
 weight percent, 14 KCl, 28 SnCl2 , and 58 
 PbCl2, The electrolyte melting point is 
 390° C, Refined tin-lead solder is re- 
 covered at the cathode. In tests, gold 
 at the anode increased from about 60 to 
 2,200 tr oz/ St. The impure gold bullion 
 collected at the anode is treated by con- 
 ventional fire refining to recover rela- 
 tively pure gold. 
 
 A second method for recycling spent 
 tin-lead solder is by drossing with alu- 
 minum or zinc (13). The electronic sol- 
 ders used in studying this method were 
 nominal 60-40 tin-lead solders that had 
 become so contaminated during wave sol- 
 dering of printed circuit boards that 
 they would no longer form acceptable 
 bonds. The analyses of two contaminated 
 solders used in the drossing study are 
 listed in table 17. 
 
 TABLE 17. - Analyses of some 
 as-received scrap electronic 
 solders, parts per million 
 
 Recovery of gold from electronic sol- 
 ders by phase separation requires a den- 
 sity difference between the solder and 
 the drossing agent. Scrap electronic 
 solders have specific gravities ranging 
 from 8.45 to 8.85 g/cm^ , while the spe- 
 cific gravities of aluminum and zinc are 
 2.70 and 7.14 g/cm^ , respectively. 
 
 Samples of scrap solder were melted in 
 open clay and clay-graphite crucibles in 
 a conventional pot furnace. Bath temper- 
 atures were held at 550° C for aluminum- 
 treated solders and at 350° C for zinc- 
 treated solders. Agitation of the bath 
 during cooling helped phase separation. 
 When the bath temperature reached 200° to 
 250° C, the dross was removed from the 
 solder by filtering. 
 
 In addition to gold, the aluminum also 
 removes significant amounts of antimony, 
 copper, iron, and nickel from the solder, 
 but is not effective for silver removal. 
 Zinc is as effective as aluminum for re- 
 moving gold from solder and also removes 
 silver, copper, iron, and nickel, but not 
 antimony. The use of zinc, however, im- 
 poses distinct disadvantages. Compared 
 with aluminum, approximately eight times 
 as much zinc, on a molar basis, is needed 
 to completely remove gold from the sol- 
 der. Zinc dross does not become as vis- 
 cous as aluminum dross as the bath tem- 
 perature is lowered, thus complicating 
 phase separation. In addition, solder 
 dissolved larger quantities of zinc than 
 aluminum, making it more difficult to re- 
 purify the solder to an acceptable elec- 
 tronic grade. Therefore, in general, 
 aluminum drossing is preferred over zinc 
 drossing except for solders containing 
 substantial quantities of silver. 
 
 Dross recovered by filtration contains 
 essentially all of the gold originally 
 present in the solder and typically rep- 
 resents about 10 pet of the weight of the 
 solder. This dross can be further pro- 
 cessed to concentrate the gold, since 
 about 85 pet of the dross is entrained 
 solder. 
 
55 
 
 Heating the dross in the range of 800° 
 to 1,000° C will oxidize aluminum and 
 aluminum-gold compounds together with 
 some of the lead and tin. Most of the 
 metallic solder entrained in the dross, 
 amounting to 15 to 40 wt pet of the 
 dross, remains in the metallic state and 
 is separated from the oxidized material 
 by grinding and screening. The metallic 
 portion remains on a 200-mesh screen, 
 whereas the oxidized material passes 
 through. The entrapped solder recovered 
 from the dross has approximately the same 
 concentration of gold as the original 
 scrap solder and is recycled when a fresh 
 batch is treated. 
 
 The gold in the oxide phase is in the 
 free metallic state and can be reclaimed 
 by aqua regia digestion or by a combined 
 cyanidation-amalgamation procedure. If 
 aqua regia digestion is chosen, the base 
 metal oxides should first be treated to 
 remove them so they do not report to 
 the aqua regia with the gold. This is 
 best accomplished by leaching the oxides 
 successively with a 50-pct NaOH solution 
 and concentrated HCl. The NaOH solution 
 removes most of the AI2O3, while HCl 
 dissolves the remaining soluble impuri- 
 ties. After filtering and washing, the 
 solutions are discarded. The remaining 
 
 solids are treated in two steps with 
 aqua regia, then filtered and washed. 
 Approximately 98 pet of the gold dis- 
 solves in the aqua regia. Gold can be 
 recovered from the aqua regia solution by 
 conventional cementation procedures. In 
 cemented form, the gold can be easily 
 processed by a gold smelter for final 
 purification. 
 
 The second technique utilizes combined 
 cyanidation-amalgamation. Neither cyani- 
 dation nor amalgamation alone is effec- 
 tive because of the variation in the size 
 distribution of the gold particles. The 
 larger particles are removed by amalgama- 
 tion, while the smaller particles are 
 dissolved by the cyanide solution. In 
 order to use the combined cyanidation- 
 amalgamation procedure, the oxides must 
 first be treated with 5N HNO3 . This 
 treatment is necessary to remove a tar- 
 nished surface film from the gold that 
 hinders reaction of the gold with the cy- 
 anide solution or with mercury. HNO3 
 also dissolves unoxidized solder and some 
 base metal oxides. Combined cyanidation- 
 amalgamation is generally preferred over 
 aqua regia digestion because of the high- 
 ly corrosive nature of aqua regia, which 
 complicates the selection of construction 
 materials. 
 
 CONCLUSIONS 
 
 Efforts to recover precious metals from 
 electronic and electrical scrap will 
 benefit from initial hand segregation 
 of the scrap material, in nearly every 
 case, even if the segregation is only 
 minimal. Base metals recovered will usu- 
 ally pay for the segregation step. Me- 
 chanical processing of electronic and 
 electrical scrap by companies specializ- 
 ing in this operations is available on a 
 toll basis. The material fractions ob- 
 tained from mechanical processing are 
 homogeneous enough for assaying. Assays 
 
 enable the generator or owner to deter- 
 mine whether the scrap has enough value 
 to at least pay for shipping and toll re- 
 fining charges. The Bureau's studies in 
 electronic and electrical scrap recovery 
 have shown that mechanical processing to 
 obtain homogeneous materials for assaying 
 is beneficial and economical, and a study 
 conducted by private industry has borne 
 out this conclusion. However, further 
 processing to recover precious metals is 
 best handled by toll refiners. 
 
56 
 
 REFERENCES 
 
 1. Dunning, B. W. , Jr. Characteriza- 
 tion of Scrap Electronic Equipment for 
 Resource Recovery, Paper in Proceedings 
 of the Sixth Mineral Waste Utilization 
 Sjmiposium (IIT Res. Inst., Chicago, XL, 
 May 2-3). ITT Res. Inst., 1980, 1978, 
 pp. 403-410. 
 
 2. Dunning, B. W. , Jr., and F. Am- 
 brose. Characterization of Pre-1957 Avi- 
 onic Scrap for Resource Recovery. Bu- 
 Mines RI 8499, 1980, 20 pp. 
 
 3. Ambrose, F., and B. W. Dunning, Jr. 
 Precious Metals Recovery From Electronic 
 Scrap. Proceedings of the Seventh Miner- 
 al Waste Utilization Symposivim (ITT Res. 
 Inst., Chicago, IL, Oct. 20-21, 1980). 
 ITT Res. Inst., 1980, pp. 184-197. 
 
 4. . Mechanical Processing of 
 
 Electronic Scrap To Recover Precious- 
 Metal-Bearing Concentrates. Ch. in Pre- 
 cious Metals, ed. by R. 0. McGachie and 
 
 Pergamon, 1980, pp. 67- 
 
 A. G. 
 76. 
 5. 
 
 H. V. 
 
 Bradley. 
 
 Dunning, B. W. Jr. , F. Ambrose, and 
 Makar. Distribution and Analysis 
 of Gold and Silver in Mechanically Pro- 
 cessed Mixed Electronic Scrap. BuMines 
 RI 8788, 1983, 17 pp. 
 
 6. Hilliard, H. E. , B. W. Dunning, 
 Jr. , and H. V. Makar. Hydrometallurgical 
 Treatment of Electronic Scrap Concen- 
 trates Containing Precious Metals. 
 BuMines RI 8757, 1983, 15 pp. 
 
 7. Hilliard, H. E., B. W. Dunning, 
 Jr. , D. A. Kramer, and D. M. Soboroff . 
 
 Hydrometallurgical Treatment of Electron- 
 ic Scrap To Recover Gold and Silver. 
 BuMines RI 8940, 1985, 20 pp. 
 
 8. Hilliard, H. E., and B. W. Dun- 
 ning, Jr. Recovery of Platinum-Group 
 Metals and Gold From Electronic Scrap. 
 Paper in Proceedings, 1983 International 
 Precious Metals Institute International 
 Seminar. The Platinum Group Metals — An 
 In-Depth View of the Industry, ed. by 
 D. E. Lundy and E. D. Zysk (Williamsburg, 
 VA, Apr. 10-13, 1983). Int. Precious 
 Met. Inst., 1983, pp. 129-142. 
 
 9. Salisbury, H. B., L. J. Duchene, 
 and J. H. Bilbrey, Jr. Recovery of Cop- 
 per and Associated Precious Metals From 
 Electronic Scrap. BuMines RI 8561, 1981, 
 16 pp. 
 
 10. Sullivan, T. A., R. L. deBeau- 
 champ, and E. L. Singleton. Recovery of 
 Aluminum, Base, and Precious Metals From 
 Electronic Scrap. BuMines RI 7617, 1972, 
 16 pp. 
 
 11. Dannenberg, R. 0., J. M. Maurice, 
 and G. M. Potter. Recovery of Precious 
 Metals From Electronic Scrap. BuMines RI 
 7683, 1972, 19 pp. 
 
 12. Kleespies, E. K. , J. P. Bennetts, 
 and T. A. Henrie. Gold Recovery From 
 Scrap Electronic Solders by Fused-Salt 
 Electrolysis. BuMines TPR 9, 1969, 8 pp. 
 
 13. Ferrell, E. F. Recovering Gold 
 From Scrap Electronic Solders by Dross- 
 ing. BuMines RI 8169, 1976, 9 pp. 
 
 T^U.S. GPO: 1985-605-017/20,138 
 
 INT.-BU.O F MIN ES,PGH.,P A. 28195 
 
w 
 
 
J V^^'/ \/^'^\/ "-^'^^•/ V-- 
 
 ° <c,'^'^'V, °oWot*" -^^^'^ '•^^*'' .^'^'^'\, °-w^** '^.'^^'^'^.^ ^^K*" J'''^^'^ 
 
 q.. ' * o Vo ' ^^0' "^^^ * • ' ^ ° ^V' . . ^^_ ' " « » " ^"^ _ ^ ''^^ '." ^ " ^V?-' . . . "-^ 
 
 .^^'V 
 
 v ^ 
 
 o_ o> 
 
 O^ 'o. » » A <. *■' ..5^ ^G 
 
 .0 V *; 
 
 '^^^^^^ ;^fe\ ^^Z •^^^"° ""^v-/ 
 
 
 
 
 
 ' <^ 
 
 '-^o^ 
 
 'b V" 
 
 ^v '^ o « o , . 
 
 ^. ^^ A^ ^^fe'- ^^ -^-^ 
 
 h' / "^^^ ^yW:* 4.^ "^^ 
 
 
 A <'^ '..s 
 
 
 ^' % J' iM/K^ ^^.A^ 
 
 ^^-n^^ 
 
 A^ --- "'^- 
 
 v^ .^=:;c.'* <^. 
 
 
 ^""^t. 
 
 J"^^ 
 
 v ^ 
 
 
 c> 'o,»- A <*^ *'V.»' .G^ -^ 
 
 ,•1°^ 
 
 "\/ 
 
 y^° ^\^ "...-.-.... .....-,. - . .^.., ,..^. 
 
 t:> 'o^.^o A <^ *^T^ <G^ ^^. '»-- ^<^ 
 
 
 
 *'^M7p^^^ ^ .r /^^\\i^%.^ -^^ ^. - :mi//^- ^ .^ i,^^&[- ^^Mr^ 
 
 o V 
 

 ^ 
 
 '■"'« '^o 
 
 
 
 .•i<^^ 
 
 V >*'% •.^- ./'\. °.^.- **'\ \^- /•% \^.- ^*^% --,1 
 
 •• O /in 
 
 
 o « ■ ^,0 
 
 -^..^ 
 
 ^^ "^. 
 
 « o , ,i» 
 
 
 
 r «. ' • o, V- 
 
 
 "''^_ 
 
 
 
 .0^ ^'^^ 
 
 1^ » 
 
 « ' .0 
 
 r . < • 
 
 
 Lf''^ O 
 
 •\/\ "•«^'' /\ •.^:° /\ '•fm' **'% --IK* >^'\ 
 
 
 -TAq^ 
 
 „*lo* 
 
 «*1«^^ 
 
 &^ o. 
 
 BOOKBINDING I 
 
 Crantville. Pa. 
 Mar — Apr 1986 
 
 IVeVe Q^ji.iy Bound 
 
 y ..... -^^ """'^^'^ ,. ^-^ '"^" ^^ 
 
 o ly "7*. 
 
 WERT ^^^^1^^ -^^ A-^ ^^&o % ^^ *^fe'' ''^^^ .b^" » 
 
 ' -V %> ' ^iiP * -^^ ^, oVJCW* aV -^i, 
 
 ^°-n^. 
 
 c" . 
 
 •^0^ 
 
 .X* o * 
 
 o 
 
 
 A < 
 
 
 
 •^^ c ■> "o « "^ 
 
 
 <-. ^